Tag Archives: Benchmark

Realme Responds To AnTuTu Claims Of Cheating!

Realme On AnTuTu Cheating : A Bug? No Guidelines?

Realme just called us up personally to explain their side of the story, and sent us an official response to AnTuTu’s shocking claims of benchmark cheating!

Find out what’s going on, and what this benchmark cheating allegation could mean for the Realme brand, and potential Realme GT customers!

 

Realme On AnTuTu Cheating : It Was A Bug!

After we posted our coverage of AnTuTu exposing how the new Realme GT cheats at their benchmark, we received a personal call from Realme PR.

We were told that this was simply a misunderstanding of sorts – there was “a technical bug” that caused the “scoring issue“.

Unfortunately, they have yet to reveal what bug would cause the CPU thread delay and mosaic colour block modifications of the AnTuTu Benchmark in Realme GT.

We will update you with the details of this bug, when we receive them from Realme PR.

They also claimed that “for the last few months”, they “tried not to mention anything about AnTuTu Benchmark”, or “associate Realme with AnTuTu Benchmark”.

At this point, we should share that Realme CMO Xu Qi Chase himself posted this (now) infamous screenshot of Realme GT scoring over 770,000 on AnTuTu Benchmark on 20 February 2021 – less than a month ago.

 

Realme On AnTuTu Cheating : There Were No Guidelines!

Realme then followed up with an official statement, claiming that the realme GT benchmark results are accurate.

They also claimed that the misunderstanding was due to a lack of “official benchmark standards or guidelines“.

Benchmark scores of realme GT are all accurate data under the current versions of the Antutu benchmark.

realme has always been committed to bringing users excellent performance, the actual user experience is the top priority for realme.

On the other hand, regarding the test issues doubted by both sides, we are communicating with Antutu positively now.

Since there is no official benchmark standards or guidelines now, so maybe there are some different understanding between manufacturers and Antutu on how to get a higher benchmark score, so before the final conclusion, we think “cheating” may be misleading for followers, it’s better to change it to “an unverified high score”.

What happened to the bug???

As for guidelines when it comes to benchmarking, we are pretty sure it is as simple as :

  1. Update to the latest firmware.
  2. Clear all other apps.
  3. Run the benchmark as-is, without modifications.

What other guidelines do you need to run AnTuTu, or any other mobile benchmark really?

 

Realme GT Cheating In AnTuTu Benchmark : A Quick Primer!

AnTuTu just exposed how Realme cheated at their benchmark to obtain the phenomenally high score of over 770,000 that Xu Qi Chase posted above.

After analysing the performance of a retail Realme GT unit, they discovered that it implemented two cheats :

  • Thread Delay : falsely inflating the CPU multi-thread test score by intercepting threads and assigning them to the fast cores.
  • Mosaic Colour Block Modification : falsely inflating the UX score by replacing the details in the reference JPG image with colour blocks.

They then warned that if Realme does not remove the cheats within 3 months, they will ban Realme GT permanently.

Recommended : Realme GT Faces Ban After Cheating In AnTuTu Benchmark!

 

What Does Realme GT Cheating Mean For Potential Customers?

If AnTuTu allegations are true, this is going to really spike the sale of the Realme GT, which is set to launch in many countries.

Especially when this incident reminds potential buyers of the time when Realme was caught faking the 90 Hz display on the Realme 6 and Realme 6 Pro.

Perhaps that is why Realme has been reluctant to send their smartphones to be reviewed, relying instead on influencers and online sales to sell their smartphones.

This new AnTuTu Benchmark cheating scandal could be a serious blow to Realme’s credibility, and a tipping point for the Realme brand.

It became a fast-rising brand by borrowing the Xiaomi model of offering a ton of features at very low prices.

The fact that Realme has a flagship smartphone powered by the latest Snapdragon 888 SoC should be a testament to how far it came, so quickly.

Yet it appears that they will go way beyond aggressive pricing and other usual tactics to win market share.

Potential customers (and other benchmark developers) will now wonder – if the Realme GT was caught cheating in AnTuTu, could it also be cheating in other benchmarks?

Even worse – what other Realme smartphones employ such tactics to obtain higher than normal scores? This will be a real PR headache.

We should point out though that benchmark cheating is not limited to smartphone brands. Even MediaTek was caught cheating in 8 of their chipsets, involving more than 50 smartphones from Xiaomi, OPPO, Vivo and yes, Realme too.

Recommended : MediaTek Benchmark Cheating : What Happened So Far

 

Recommended Reading

Go Back To > Mobile | Software | Home

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


Realme GT Faces Ban After Cheating In AnTuTu Benchmark!

The Realme GT faces a possible permanent ban after it was caught cheating in the AnTuTu Benchmark!

Find out what’s going on, and what it means to you – potential Realme customers!

 

Benchmark Cheating : Why Do It?

The smartphone industry is a cut-throat business, with brands fighting to differentiate their smartphones from their competitors. One of the ways is to demonstrate a clear performance advantage.

Other than the usual ways – using more powerful processor and graphics cores – brands can tune their operating system to focus more on performance, at the expense of battery life.

However, when they use the same mobile platform, a desperate brand may resort to cheating to gain an edge over their rivals.

AnTuTu is accusing Realme of doing just that with their flagship Realme GT smartphone.

 

Realme GT Caught Cheating In AnTuTu Benchmark?

On 15 March 2021, AnTuTu posted their Weibo expose on how the Realme GT cheated at their benchmark. For your convenience, we are reposting it at the end of this article with our English translation.

AnTuTu first became suspicious when Realme declared an AnTuTu Benchmark result of over 770,000 for their Realme GT smartphone before it launched.

When Realme did not respond to their questions, AnTuTu purchased a unit of the Realme GT to test it for themselves.

After analysing the performance of their unit, they discovered that the Realme GT implemented two cheats :

Thread Delay In CPU Multi-Thread Test

The Realme GT was allegedly designed to intercept the threads from the CPU Multi-Thread test, to schedule them to run on the fast cores as much as possible.

This falsely inflates its score by using the fast cores to process the threads, instead of testing the processor’s true multi-core capability in handling multiple threads concurrently.

Mosaic Colour Block Modification In UX Test

AnTuTu discovered that the Realme GT modified the reference JPG image with mosaic colour blocks.

By replacing the details in the original picture with blocking mosaics, the JPG decoding speed can be greatly inflated.

 

Realme GT Threatened With Permanent Ban

After announcing their results, AnTuTu announced that they removed all Realme GT scores for 3 months.

They also warned that if Realme does not remove the cheats within 3 months, they will ban Realme GT permanently.

AnTuTu also shared that their upcoming V9 benchmark will feature improved anti-cheating features, and even trigger penalties like score deductions.

 

Realme : AnTuTu Scores Real, Will No Longer Using AnTuTu?

Realme CMO Xu Qi Chase quickly shot back a response to AnTuTu’s expose.

If we understand this correctly, he is insisting that the Realme GT scores are accurate, and announcing that Realme will no longer use the AnTuTu Benchmark results for future Realme smartphones.

 

What Does Realme GT Cheating Mean For Potential Customers?

If AnTuTu allegations are true, this is a serious blow to Realme’s credibility.

It became a fast-rising brand by borrowing the Xiaomi model of offering a ton of features at very low prices.

The fact that Realme has a flagship smartphone powered by the latest Snapdragon 888 SoC should be a testament to how far it came, so quickly.

Yet it appears that they would go way beyond aggressive pricing and other usual tactics to win market share.

Potential customers (and other benchmark developers) will now wonder – if the Realme GT was caught cheating in AnTuTu, could it also be cheating in other benchmarks?

Even worse – what other Realme smartphones employ such tactics to obtain higher than normal scores? This will be a real PR headache.

We should point out though that benchmark cheating is not limited to smartphone brands. Even MediaTek was caught cheating in 8 of their chipsets, involving more than 50 smartphones from Xiaomi, OPPO, Vivo and yes, Realme too.

Recommended : MediaTek Benchmark Cheating : What Happened So Far

 

AnTuTu Expose Of Realme GT Cheating

Here was what AnTuTu posted in their Weibo post, with our English translation below :

realme真我GT上市前,官方曾公布过高达76万以上的安兔兔测试成绩,对此我们也曾提出过质疑,但至今并未得到realme方面的具体回应。

当时我们也已承诺,待真我GT上市后会购入这款机型进行相关测试。今天这份下架说明,则是在我们购入量产机,并实际测试后所作出的决定。

我们所购的这台真我GT于3月10日首发当天,在欢太商城购入,其存储配置为12+256GB。

通过多次测试,这台真我GT在安兔兔V8版本中均为75万分左右的成绩,在目前在售搭载骁龙888平台的机型中,已经达到了非常出色的水准。但根据我们在对其测试过程中所进行的数据分析发现,这并非真实实力的体现,而是通过作弊等手段所取得。

通过我们的验证,真我GT在CPU多线程测试中存在线程延迟的情况,通过系统拦截线程的方案,将多线程测试的部分线程尽可能调度到大核上运行,导致测试成绩更高。

这种做法已经背离了安兔兔测试的初衷,多线程测试所考察的是多线程并发的处理器多核性能,并不是为了考察大核在连续运行测试时候的性能表现。

另外在UX测试的JPG解码项目中,我们发现真我GT跑分时将安兔兔测试的解码图片进行了修改,出现了大量马赛克。使用马赛克色块替换原有图片细节可以显著降低图片质量,从而解码速度也会更快,所以才有了更高的分数。但这样做明显违背了图片解码测试的目的。

综上所述,安兔兔方面决定下架realme真我GT的跑分成绩三个月,如未来realme方面能够修改上述问题,三个月后会重新计入相关榜单,如拒不修改则将其永久下架。

在即将到来的V9版本中,安兔兔将进一步优化防作弊方案,例如出现作弊情况会引发扣分等一些处罚方式。

Before the realme GT went on the market, the official announcement of the Antutu test results of more than 760,000. We questioned this, but so far we have not received a specific response from realme.

At that time, we also promised that we will purchase this model to test after the Realme GT launched. Today’s removal instructions were made after we purchased the mass production device and actually tested it.

The Realme GT we bought was purchased on the Huantai Mall on the day of its debut on March 10th, and its storage configuration is 12+256GB.

After many tests, this realme GT scored about 750,000 points in the AnTuTu V8 version, and it has reached a very good level among the models currently on sale with the Snapdragon 888 platform. But according to the data analysis we conducted during the testing process, we found that this is not a manifestation of true strength, but obtained through cheating and other means.

Through our verification, Realme GT used thread delays in the CPU multi-thread test. Through the system interception of threads, some threads of the multi-thread test are scheduled to run on large cores as much as possible, resulting in higher test results.

This approach has deviated from the original intention of the AnTuTu test. The multi-threaded test examines the multi-threaded concurrent processor multi-core performance, not to examine the performance of the large core during continuous running tests.

In addition, in the JPG decoding project of the UX test, we found that the decoded pictures of the Antutu test were modified during the real GT running, and a lot of mosaics appeared. Using mosaic color blocks to replace the original picture details can significantly reduce the quality of the picture, and thus the decoding speed will be faster, so it has a higher score. But doing so obviously violates the purpose of the picture decoding test.

To sum up, AnTuTu decided to remove the realme GT’s running scores for three months. If realme can modify the two behaviours, it will be included in the relevant list again in three months. If it refuses to modify it, it will. be banned permanently.

In the upcoming V9 version, AnTuTu will further optimize the anti-cheating program. For example, cheating will trigger some penalties such as deductions.

 

Recommended Reading

Go Back To > Mobile | Software | Home

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


Radeon RX 6000 vs GeForce RTX 30 : Faster + Cheaper!

AMD claims that their Radeon RX 6000 graphics cards are faster and cheaper than the GeForce RTX 30 graphics cards NVIDIA recently launched.

Take a look at the performance results comparing their Radeon RX 6000 cards against the NVIDIA GeForce RTX 30 cards!

 

Radeon RX 6000 vs GeForce RTX 30 : Faster + Cheaper!

The AMD Radeon RX 6000 Series is built upon the new RDNA 2 architecture, which introduces an enhanced compute unit, a new visual pipeline, and the all-new AMD Infinity Cache!

RDNA 2 promises to deliver up to 54% better performance per watt over the last-generation RDNA, and up to 3.25X more bandwidth with its new Infinity Cache.

In this video, Dr. Lisa Su, Scott Kerkelman and Laura Smith shares how the new Radeon RX 6000 cards beat NVIDIA’s GeForce RTX 30 cards in performance, power efficiency and yes, price!

 

Radeon RX 6000 vs GeForce RTX 30 : Performance Comparison

Caveats

  • These results were provided by AMD, and have yet to be verified by independent reviewers / end users.
  • The games were tested with different APIs (DX12, Vulkan) and the best results were used. They are therefore, not directly comparable.
  • The AMD Radeon RX 6900 XT and RX 6800 cards were tested with Rage Mode and/or Smart Access Memory enabled, which may not reflect real world performance.

Radeon RX 6900 XT vs GeForce RTX 3090 : 4K Gaming Performance

According to AMD, their Radeon RX 6900 XT will match or beat the GeForce RTX 3090, at a much lower price point and significantly lower power consumption.

Radeon RX 6800 XT vs GeForce RTX 3080 : 4K Gaming Performance

At 4K, AMD promises that their Radeon RX 6800 XT will match or beat the GeForce RTX 3080, at a lower price point and with a lower power consumption.

Radeon RX 6800 XT vs GeForce RTX 3080 : 1440p Gaming Performance

At 1440p, the Radeon RX 6800 XT is roughly equal to GeForce RTX 3080 in performance, at a lower price point and with a lower power consumption.

Radeon RX 6800 vs GeForce RTX 2080 Ti : 4K Gaming Performance

Depending on the game, the Radeon RX 6800 will match or be significantly faster than the GeForce RTX 2080 Ti, at a significantly lower price point.

Radeon RX 6800 vs GeForce RTX 2080 Ti : 1440p Gaming Performance

At 1440p, the Radeon RX 6800 is significantly faster than the GeForce RTX 2080 Ti, at a significantly lower price point.

 

Recommended Reading

Go Back To > Computer HardwareHome

 

Support Tech ARP!

If you like our work, you can help support us by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


Snapdragon vs Exynos vs Kirin : Battle Of The Flagships!

Your fancy new flagship smartphone is probably powered by a Snapdragon, Exynos or Kirin mobile platform, but just how fast are they really?

Find out which is the fastest flagship mobile SoC in our little performance comparison!

 

Snapdragon vs Exynos vs Kirin : Battle Of The Flagships!

So we managed to get our hands on the fastest mobile platforms available in the market today :

Snapdragon
865+
Exynos
990
Kirin
990
Process 7 nm N7P 7 nm EUV 7 nm EUV
CPU
Cores
1 x Kryo 585 Gold
(3.09 GHz)
3 x Kryo 585 Gold
(2.42 GHz)
4 x Kryo 585 Silver
(1.8 GHz)
2 x Exynos M5
(2.7 GHz)
2 x Cortex-A76
(2.5 GHz)
4 x Cortex-A55
(2.0 GHz)
2 x Cortex-A76
(2.86 GHz)
2 x Cortex-A76
(2.09 GHz)
4 x Cortex-A55
(1.8 GHz)
GPU Adreno 650 Mali-G77 MP11 Mali-G76 MC16
Memory
Support
LPDDR5 LPDDR5 LPDDR5
Storage
Support
UFS 3.0 UFS 3.0 UFS 3.0
5G Modem Snapdragon X55
(Separate)
Exynos 5123
(Separate)
Balong 5000
(Integrated)

Now you may be wondering – what about MediaTek? They get a pass because they cheat at benchmarks.

 

Snapdragon vs Exynos vs Kirin : TLDR Summary

For those who are too lazy to drill down into the results, here’s a quick and dirty summary.

The Qualcomm Snapdragon 865 Plus is definitely the superior mobile platform.

In most benchmarks, it was significantly faster than both the Samsung Exynos 990 and the HUAWEI Kirin 990.

Snapdragon
865+
Exynos
990
Kirin
990
PCMark Work 2.0 Winner -16.3% -30.6%
Computer Vision Winner -16.9% -23.2%
3DMark Vulkan Winner -11.4% -33.7%
3DMark OpenGL Winner -4.5% -30.1%
Geekbench Multi-Core Winner -3.2% -3.7%
Geekbench Single-Core -21.1% -2.2% Winner
AnTuTu Overall Winner -9.5% -22.8%
Average Winner -6.1% -21.4%

You are unlikely to notice the small 6% performance difference between the Snapdragon 865+ and the Exynos 990.

However, the 21% performance deficit of the HUAWEI Kirin 990 versus the Snapdragon 865+ will be noticeable. So will its 14% performance deficit against the Exynos 990.

Even if TSMC was allowed to continue making Kirin 990 for HUAWEI, it is no longer a competitive flagship mobile platform.

 

Snapdragon vs Exynos vs Kirin : PCMark Work 2.0

Snapdragon 865+ Exynos 990 Kirin 990
Work 2.0 Score Winner -16.3% -30.6%
Web Browsing 2.0 Winner -27.2% -43.3%
Video Editing Winner -18.5% -26.2%
Writing 2.0 Winner -12.0% -13.4%
Photo Editing 2.0 -10.7% Winner -45.6%
Data Manipulation Winner -29.6% -26.9%

 

Snapdragon vs Exynos vs Kirin : PCMark Computer Vision

This is a benchmark that tests the AI capabilities of these mobile platforms :

  • image recognition with TensorFLow
  • graphical barcode scanning with ZXing
  • optical character recognition with Tesseract

Snapdragon 865+ Exynos 990 Kirin 990
Computer Vision Winner -16.9% -23.2%
TensorFlow -44.2% -5.0% Winner
ZXing Winner -18.1% -3.6%
Tesseract -19.9% Winner -56.6%

Next Page > Snapdragon vs Exynos vs Kirin : 3DMark | Geekbench | AnTuTu

 

Support Tech ARP!

If you like our work, you can help support us by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


Snapdragon vs Exynos vs Kirin : 3DMark Sling Shot Extreme

We tested them using the Sling Shot Extreme test in 3DMark, using both Vulkan and OpenGL ES 3.1 graphics API.

Vulkan is a newer graphics API, with much lower overhead, and can therefore deliver better performance than OpenGL ES which is supported by more games.

3DMark rendered the game at the 2560 x 1440 resolution, allowing for a fair comparison irrespective of the devices’ actual display resolution.

Snapdragon 865+ Exynos 990 Kirin 990
Vulkan Overall Winner -11.4% -33.7%
Vulkan Graphics Winner -17.0% -48.8%
Vulkan Physics -6.3% -9.1% Winner
OpenGL Overall Winner -4.5% -30.1%
OpenGL Graphics Winner -6.2% -41.3%
OpenGL Physics -2.4% -4.0% Winner

 

Snapdragon vs Exynos vs Kirin : Geekbench

Snapdragon 865+ Exynos 990 Kirin 990
Multi-Core (MC) Winner -3.2% -3.7%
MC Crypto -27.2% -23.9% Winner
MC Integer Winner -4.2% -6.0%
MC Floating Point Winner -2.9% -7.7%
Single-Core (1C) -21.1% -2.2% Winner
1C Crypto -33.2% -10.3% Winner
1C Integer -19.4% -3.7% Winner
1C Floating Point -23.8% Winner -3.1%

 

Snapdragon vs Exynos vs Kirin : AnTuTu

Snapdragon 865+ Exynos 990 Kirin 990
Overall Winner -9.5% -22.8%
CPU Winner -9.0% -15.7%
GPU Winner -6.8% -37.6%
MEM -2.6% -11.5% Winner
UX Winner -18.3% -25.5%

 

Recommended Reading

Go Back To > First PageMobile Devices | Home

 

Support Tech ARP!

If you like our work, you can help support us by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


Samsung Galaxy Note 20 Series : Exynos 990 Performance!

Take a look at the work and gaming performance of the Exynos 990 version of the Samsung Galaxy Note 20 series smartphones!

 

Samsung Galaxy Note 20 Series : Powered By Exynos 990

Globally, the Samsung Galaxy Note 20 and Note 20 Ultra smartphones are powered by the Samsung Exynos 990, the same mobile SoC used by the Samsung Galaxy S20 smartphones.

Faster Processor + GPU

Its custom Mongoose 5 (Exynos M5) cores are 20% faster than the Mongoose 4 (Exynos M4) cores used in the flagship-class Exynos 9825 that powered last year’s Galaxy Note 10 smartphones.

Its Arm Mali-G77 GPU uses the new Valhall architecture, which promises to offer a 20% boost in graphics performance or power efficiency.

Faster LPDDR5 Memory

The Exynos 990 is the first Samsung mobile platform to use the new LPDDR5 memory, which offers a 31% higher clock speed – increased from 2093 MHz to 2750 MHz.

This allows it to deliver up to 5,500 Mbps or 687.5 MB/s in memory bandwidth. It should also allow for lower power consumption.

Faster NPU + DSP

Its AI capabilities also received a big boost with a new dual-core NPU and improved digital signal processor (DSP) that can process over 10 trillion operations (TOPS) per second!

Recommended : Samsung Exynos 990 : Everything You Need To Know!

 

Samsung Galaxy Note 20 Series : Exynos 990 Performance!

Note : We tested the Galaxy Note 20 Ultra against last year’s Galaxy Note10+ and Galaxy S10+ devices, as well as the two fastest Qualcomm devices we have in the lab, including the latest Qualcomm Snapdragon 865+.

However, please note that the Snapdragon 865+ results are from the Galaxy Tab S7+ tablet, not a smartphone.

Work Performance : PCMark

We tested the Galaxy Note 20 Ultra using PCMark, which simulates work applications like web browsing, playing video, writing text and editing photos.

Work 2.0 Galaxy
Tab
S7+
Galaxy
Note20
Ultra
Galaxy
Note10+
Black Shark
2 Pro
Galaxy
S10+
Mobile Platform SD 865+ Exynos 990 Exynos 9825 SD 855+ Exynos 9820
Performance Score
(Average)
12348 10523 8981 8384 7846
Web Browsing 11864 8865 8293 6768 8583
Video Editing 7664 6232 9166 5948 5524
Writing 12659 11112 9579 9226 9076
Photo Editing 26230 30789 19376 17595 10710
Data Manipulation 9511 6825 6155 6339 6450

Samsung Exynos has historically been rather poor at work performance. But the Exynos 990 was significantly faster than the last-generation Exynos 9825 and Exynos 9820.

Thanks to the improved performance, the Galaxy Note 20 Ultra was 17% faster than last year’s Galaxy Note10+ (Exynos 9825), and 25% faster than the Black Shark 2 Pro, which is powered by Snapdragon 855+.

However, the Snapdragon 865+ is 17% faster than the Exynos 990 in this test, losing out only in the Photo Editing test (by 15%).

Gaming Performance : 3DMark

For gaming, we first tested the Galaxy Note 20 Ultra using 3DMark, with these results :

3DMark Galaxy
Tab
S7+
Galaxy
Note20
Ultra
Black Shark
2 Pro
Galaxy
Note10+
Galaxy
S10+
Device SD 865+ Exynos 990 SD 855+ Exynos 9825 Exynos 9820
Sling Shot Extreme (OpenGL) 6958 6642 6258 5037 4470
Sling Shot Extreme (Vulkan) 6457 5722 5386 4794 4314
Sling Shot 8793 8128 8127 4883 4343

The Samsung Galaxy Note 20 Ultra was as fast as the Black Shark 2 Pro (Snapdragon 855+) in Sling Shot, and 6% faster in Sling Shot Extreme.

However, the Exynos 990 was no match for the Snapdragon 865+, which was 5% faster (OpenGL) to 13% faster (Vulkan) in Sling Shot Extreme.

Gaming Performance : AnTuTu

We then tested the Galaxy Note 20 Ultra using AnTuTu, yielding these results :

AnTuTu Galaxy
Tab
S7+
Galaxy
Note20
Ultra
Black Shark
2 Pro
Galaxy
Note10+
Galaxy
S10+
Device SD 865+ Exynos 990 SD 855+ Exynos 9825 Exynos 9820
AnTuTu Score 577726 522735 434168 348397 330549
CPU 161243 146783 126170 101187 100272
GPU 230859 215205 188158 156700 149930
Memory 98269 89376 65052 72051 68188
UX 87355 71371 54788 18459 12159

In AnTuTu, the Samsung Galaxy Note 20 Ultra was 20% faster than the Black Shark 2 Pro (Snapdragon 855+) and 40% faster than the Xiaomi Mi 9 (Snapdragon 855)!

The Galaxy Note 20 Ultra was even faster when we compared it to last year’s Galaxy Note10+ (Exynos 9825) – 50% faster, and Galaxy S10+ (Exynos 9820) – 58% faster!

But the Exynos 990 was, yet again, slower than the Snapdragon 865+ in this test. The Snapdragon 856+ was faster by 10.5%.

 

Samsung Galaxy Note 20 Series : Specifications (Exynos Version)

Specifications Galaxy Note 20 Galaxy Note20 Ultra
Display 6.7-inch Infinity-O Display
Super AMOLED+ Panel
6.9-inch Infinity-O Display
Dynamic AMOLED 2X Panel
Display Features HDR10+, 60 Hz HDR10+, 120 Hz, 1500 nits
Display Resolution 1080 x 2400 pixels 1440 x 3088 pixels
Protection Corning Gorilla Glass Victus, IP68
Operating System One UI 2.0 (Android Q)
Platform Samsung Exynos 990
Processor 2 x Mongoose 5 cores (2.7 GHz)
2 x Cortex-A76 cores (2.5 GHz)
4 x Cortex-A55 cores (2.0 GHz)
GPU ARM Mali-G77 MP11
Memory 8 GB LPDDR5 12 GB LPDDR5
Internal Storage 256 GB 256 GB
External Storage Hybrid MicroSD Slot (up to 512 GB)
Front Camera 10 MP camera (f/2.2)
Main Cameras 12 MP main camera (f/1.8)
12 MP ultra-wide camera (f/2.2)
64 MP 3X Hybrid Zoom camera (f/2.0)
108 MP main camera (f/1.8)
12 MP ultra-wide camera (f/2.2)
12 MP 5X Optical Zoom camera (f/3.0)
Laser focus sensor
Connectivity Bluetooth 5.1
WiFi-AX
USB 2.0 Type C
S Pen Yes, 9 ms response time, 4096 levels
Lithium Titanate battery (24 hours standby)
Fingerprint Sensor 3D Sonic Max
Battery 4,300 mAh 4,500 mAh
Charging 25 W 25 W
Dimensions 75.2 mm wide
161.6 mm tall
8.3 mm thick
77.2 mm wide
164.8 mm tall
8.1 mm thick
Weight 192 g 208 g

 

Samsung Galaxy Note 20 Series : Where To Buy?

Malaysia (Exynos 990)

Singapore (Exynos 990)

United States (Snapdragon 865+)

United Kingdom (Exynos 990)

Australia (Exynos 990)

 

Recommended Reading

Go Back To > Mobile Devices | Home

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


MediaTek Benchmark Cheating : What Happened So Far

MediaTek was caught cheating at smartphone benchmarks, in a stunning expose by Andrei Frumusanu from Anandtech.

Here is what you need to know about the MediaTek benchmark cheating scandal, and what it means to you – the user!

Updated 2020-04-17 : Added UL’s delisting of MediaTek chipsets, and their comments on MediaTek’s actions.

Originally posted 2020-04-09

 

Benchmark Cheating : What Is It?

The smartphone industry is a cut-throat business, with brands fighting to differentiate their smartphones from their competitors. One of the ways is to demonstrate a clear performance advantage.

Other than the usual ways – using more powerful processor and graphics cores – brands can tune their operating system to focus more on performance, at the expense of battery life.

Some though want to have the best of both worlds – the ability to brag about both high performance, as well as a long battery life.

To do that, unscrupulous brands can cheat by boosting performance only when a benchmark is run. But because this high performance mode doesn’t kick in when other apps are being used, this tricks reviewers into proclaiming that the particular device has great performance as well as great battery life.

 

Mediatek : Who Are They?

MediaTek is one of the main suppliers of mobile SoCs – the “processors” that power smartphones, tablets and even smart TVs.

They compete against Qualcomm, as a third-party supplier of these system-on-chip platforms to smartphone brands like OPPO, Vivo, and many smaller smartphone brands.

 

MediaTek Benchmark Cheating : What Did They Do?

According to Anandtech, MediaTek programmed their mobile SoCs to identify many benchmarks, and artificially boost performance, when those benchmarks run.

Hidden in the power_whitelist_cfg.xml configuration file were parameters designed to kick MediaTek-powered smartphones into a special high performance Sports Mode, when benchmarks were detected :

Credit : Anandtech

 

MediaTek Benchmark Cheating : How Much Of An Effect

Using an anonymised version of PCMark, Anandtech discovered that the cheat increased performance by 33% overall, with up to 75% boost in certain subtests like writing workload :

Work 2.0 MediaTek P95
(Real Performance)
MediaTek P95
(Sports Mode)
Difference
Performance Score
(Average)
6781 9048 +33.4%
Web Browsing 5484 9011 +64.3%
Video Editing 5306 5950 +12.1%
Writing 5734 10034 +75.0%
Photo Editing 14896 18804 +26.2%
Data Manipulation 5767 5995 +4.0%

 

MediaTek Benchmark Cheating : What Devices Are Affected?

Anandtech only checked a number of devices, but it looks like pretty much every MediaTek-powered smartphone manufactured in the last few years is affected, even the Sony XA1 from 2016!

Brand OPPO OPPO OPPO OPPO Vivo Xiaomi Realme iVoomi Sony
Device Reno3 Pro Reno Z F15 F9 Pro S1 Note 8 Pro C3 i2 Lite XA1
MediaTek SoC P95 P90 P70 P60 P65 G90 G70 A22 P20
AndroBench2 Yes Yes Yes Yes Yes Yes Yes Yes Yes
PCMark Yes Yes Yes Yes Yes Yes Yes Yes Yes
AnTuTu Yes Yes Yes Yes Yes Yes Yes Yes Yes
GeekBench Yes Yes Yes Yes Yes Yes Yes Yes Yes
Quadrant Yes Yes Yes Yes Yes Yes Yes Yes Yes
Quadrant Professional Yes Yes Yes Yes Yes Yes Yes Yes Yes
鲁大师 / Master Lu Yes Yes Yes Yes Yes Yes Yes Yes No
鲁大师 / AIMark Yes Yes Yes Yes Yes Yes Yes No No
AI Benchmark (ZTH) Yes Yes Yes Yes Yes Yes Yes No No
NeuralScope Benchmark Yes Yes Yes Yes Yes Yes Yes No No
GFXBench 4 Corporate Yes Yes No No Yes Yes Yes No No

 

MediaTek Response To Cheating Claims

After Anandtech contacted them, MediaTek issued this statement “explaining” their decision to selectively boost performance for benchmarks :

MediaTek follows accepted industry standards and is confident that benchmarking tests accurately represent the capabilities of our chipsets. We work closely with global device makers when it comes to testing and benchmarking devices powered by our chipsets, but ultimately brands have the flexibility to configure their own devices as they see fit. Many companies design devices to run on the highest possible performance levels when benchmarking tests are running in order to show the full capabilities of the chipset. This reveals what the upper end of performance capabilities are on any given chipset.

Of course, in real world scenarios there are a multitude of factors that will determine how chipsets perform. MediaTek’s chipsets are designed to optimize power and performance to provide the best user experience possible while maximizing battery life. If someone is running a compute-intensive program like a demanding game, the chipset will intelligently adapt to computing patterns to deliver sustained performance. This means that a user will see different levels of performance from different apps as the chipset dynamically manages the CPU, GPU and memory resources according to the power and performance that is required for a great user experience. Additionally, some brands have different types of modes turned on in different regions so device performance can vary based on regional market requirements.

We believe that showcasing the full capabilities of a chipset in benchmarking tests is in line with the practices of other companies and gives consumers an accurate picture of device performance.

 

UL Delists Eight MediaTek Chipsets

On 15 April 2020, UL temporarily delisted PCMark for Android scores for eight MediaTek chipsets.

  • MediaTek Helio G90
  • MediaTek Helio G70
  • MediaTek Helio P95
  • MediaTek Helio P90
  • MediaTek Helio P65
  • MediaTek Helio P60
  • MediaTek Helio P20
  • MediaTek Helio A22

This affected more than 50 smartphones from 25 smartphone brands, including the Xiaomi Redmi Note 8 Pro, Oppo Reno3 Pro, Vivo Y19, and Realme 6.

UL also called out MediaTek for claiming that they were following “industry standards”, saying :

Using hidden mechanisms to detect benchmarking apps by name and make app-specific performance optimizations is not an “accepted industry standard.” It is, in fact, the very opposite of the accepted standard.

Likewise, benchmark scores based on hidden app-specific optimizations and settings that are enabled by default and not available to the user do not accurately reflect a device’s true performance in everyday use. 

Simply put, a device must run a benchmark as if it was any other application. Performance gains must come from reacting to the nature of the workloads in the test rather than the name of the app itself.

 

MediaTek Benchmark Cheating : Conclusion

MediaTek was caught RED HANDED in the act of cheating at benchmarks. There are no two ways about it.

Their “explanation” that their chipsets are designed to “intelligently adapt to computing pattern” is just a lame excuse.

It cannot explain how their P95 processor delivered much poorer performance on an anonymised PCMark benchmark.

This is backed up by details in the configuration files that explicitly show them triggering a special Sports Mode for benchmarks.

Now, this does not change the performance of your MediaTek-powered smartphones. It just means that their benchmark scores were artificially inflated.

Short of an outright apology and a public promise never to do this again, all benchmark scores of MediaTek-powered smartphones should be considered suspect, and not to be trusted.

 

Recommended Reading

Go Back To > Mobile Devices | Home

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


AMD Radeon RX 580 Performance Comparison

The gaming enthusiast community may be eagerly waiting for AMD Vega to arrive, but AMD breathed life into the AMD Polaris microarchitecture with a refreshed line-up, which they called the AMD Radeon RX 500 Series. Headlined by the AMD Radeon RX 580 (Amazon), the Polaris Refined series is basically an overclocked version of the Radeon RX 400 series.

AMD is pitching the Radeon RX 500 Series as a great upgrade option for gamers with 3 years or older systems, with the introduction of 27 new graphics cards. But just how much faster is the new Radeon RX 580 (Amazon) compared to its predecessor, the Radeon RX 480, and the NVIDIA GeForce graphics cards? Let’s find out in this performance comparison!

[adrotate group=”1″]

 

Radeon RX 580 Specification Comparison

Here are the specifications of the five graphics cards in this performance comparison :

SpecificationsGeForce GTX 1050 TiRadeon RX 480Radeon RX 580GeForce GTX 1060GeForce GTX 1070
Texture Units4814414480120
ROPs3232324864
Core Speed1290 MHz1120 MHz1257 MHz1506 MHz1506 MHz
Boost Speed1392 MHz1266 MHz1340 MHz1708 MHz1683 MHz
Texture Fill Rate (Max)61.9 GT/s182.3 GT/s193.0 GT/s136.6 GT/s202.0 GT/s
Pixel Fill Rate (Max)44.5 GP/s40.5 GP/s42.9 GP/s82.0 GP/s107.7 GP/s
Memory Bus Width128-bits256-bits256-bits192-bits256-bits
Graphics Memory4 GB GDDR58 GB GDDR58 GB GDDR56 GB GDDR58 GB GDDR5
Memory Speed1752 MHz1750 MHz2000 MHz2000 MHz2000 MHz
Memory Bandwidth112 GB/s224 GB/s256 GB/s192 GB/s256 GB/s
TDP75 W150 W185 W120 W150 W

 

Further Reading On The Radeon RX 500 Series

 

Further Reading On AMD Vega

Next Page > Benchmarking Notes, 3DMark Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

Benchmarking Notes

Our graphics benchmarking test bed has the following specifications :

Operating System : Microsoft Windows 10 64-bit

Processor : AMD Ryzen 7 1800X processor running at 3.6 GHz

Motherboard : AORUS AX370-Gaming 5

Memory : 16 GB Corsair Vengeance LPX DDR4-3000 memory (dual-channel)

Storage : 240 GB HyperX Savage SSD

Monitor : Dell P2415Q Ultra HD Monitor

We used the GeForce driver 382.05 for the NVIDIA graphics cards, and Radeon Software 17.5.1 for the AMD graphics cards.

Note that we do not actually have any Radeon RX 580 (Amazon) graphics card. We simulated the Radeon RX 580 by overclocking the Radeon RX 480 graphics card according to AMD’s specifications for the Radeon RX 580.

[adrotate group=”1″]

 

3DMark DirectX 12 Benchmark (2560 x 1440)

3DMark Time Spy is the DirectX 12 benchmark in 3DMark. It supports new API features like asynchronous compute, explicit multi-adapter, and multi-threading.

The AMD Radeon RX 580 (Amazon) was just 1% faster than the Radeon RX 480 in this DirectX 12 benchmark. It was 69% to 76% faster than the GeForce GTX 1050 Ti, and just 3% slower than the GeForce GTX 1060.

 

3DMark (1920 x 1080)

For Direct 11 performance, we started testing the graphics cards using 3DMark at the entry-level gaming resolution – 1920 x 1080.

At this resolution, the AMD Radeon RX 580 (Amazon) was virtually neck-to-neck with the GeForce GTX 1060. It was 62% to 69% faster than the GeForce GTX 1050 Ti.

 

3DMark (2560 x 1440)

We then took 3DMark up a notch to the resolution of 2560 x 1440. Let’s take a look at the results!

At this higher resolution, the AMD Radeon RX 580 (Amazon) was about 1.5% faster than the Radeon RX 480. With its much higher pixel fillrate, the GeForce GTX 1060 pulled away with a 6% average performance advantage.

 

3DMark (3840 x 2160)

This is torture, even for the GeForce GTX 1070!

At this resolution, the GeForce GTX 1060 was still 5% to 7% faster than the Radeon RX 580 (Amazon), which was itself 61% to 65% faster than the GeForce GTX 1050 Ti.

Next Page > Ashes of the Singularity & Total War: Warhammer Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

Ashes of the Singularity (1920 x 1080)

We tested Ashes of the Singularity in the DirectX 12 mode, which supports the Asynchronous Compute feature. We started with the full HD resolution.

The AMD Radeon RX 580 (Amazon) was just 1% faster than the Radeon RX 480, and 3.7% slower than the GeForce GTX 1060.

 

Ashes of the Singularity (2560 x 1440)

We then took Ashes of the Singularity up a notch to the resolution of 2560 x 1440. Let’s see how the cards fare now…

At this higher resolution, the AMD Radeon RX 580 (Amazon) was 1.5% faster than the Radeon RX 480, and 2.3% slower than the GeForce GTX 1060. It was also 63% to 70% faster than the GeForce GTX 1050 Ti.

 

Ashes of the Singularity (3840 x 2160)

Finally, let’s see how the cards perform with Ashes of the Singularity running at the Ultra HD resolution of 3840 x 2160.

The AMD Radeon RX 580 (Amazon) was virtually tied with the Radeon RX 480 and GeForce GTX 1060 at this resolution. They were all about 67% faster than the GeForce GTX 1050 Ti, and 26% slower than the GeForce GTX 1070.

[adrotate group=”1″]

 

Warhammer (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by Total War : Warhammer‘s internal DirectX 12 benchmark.

At this resolution, the AMD Radeon RX 580 (Amazon) was 1.5% faster than the Radeon RX 480, 6.5% faster than the GeForce GTX 1070, 13% faster than the GeForce GTX 1060, and 80% faster than the GeForce GTX 1050 Ti.

 

Warhammer (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by Total War : Warhammer‘s internal DirectX 12 benchmark.

When the resolution increased to 1440p though, the GeForce GTX 1070 pulled away and was 24% faster than the Radeon RX 580. The average frame rate of the Radeon RX 580 (Amazon) was just 1.5% faster than the Radeon RX 480, but it also delivered a significantly higher minimum frame rate.

 

Warhammer (3840 x 2160)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by Total War : Warhammer‘s internal DirectX 12 benchmark.

At this ultra high resolution, the AMD Radeon RX 580 (Amazon) was now 3% faster than the Radeon RX 480. It was also 5% and 82% faster than the GeForce GTX 1060 and the GeForce GTX 1050 Ti respectively.

Next Page > The Witcher 3 & For Honor Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

The Witcher 3 (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in The Witcher 3.

The AMD Radeon RX 580 (Amazon) was 4% faster than the Radeon RX 480 at this resolution, putting it neck-to-neck with the GeForce GTX 1060.

 

The Witcher 3 (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in The Witcher 3.

At this higher resolution, the AMD Radeon RX 580 (Amazon) was essentially between the Radeon RX 480 and the GeForce GTX 1060 in performance, with a 4% difference either way. It was also 78% faster than the GeForce GTX 1050 Ti, but 27% slower than the GeForce GTX 1070.

 

The Witcher 3 (3840 x 2160)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in The Witcher 3.

 

The 4K resolution in The Witcher 3 is really tough on graphics cards, virtually halving their frame rates. The AMD Radeon RX 580 (Amazon) was, again, between the Radeon RX 480 and the GeForce GTX 1060 in performance, with a 4-5% difference either way. Only the NVIDIA GeForce GTX 1070 was able to deliver an average frame rate above 30 fps.

[adrotate group=”1″]

 

For Honor (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by For Honor‘s internal DirectX 12 benchmark.

At 1080p, the AMD Radeon RX 580 (Amazon) was exactly between the Radeon RX 480 and the GeForce GTX 1060 in performance, with a 4% difference either way. It was also 56% faster than the GeForce GTX 1050 Ti, but 31% slower than the GeForce GTX 1070.

 

For Honor (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by For Honor‘s internal DirectX 12 benchmark.

Yet again, the AMD Radeon RX 580 (Amazon) maintained its position between the Radeon RX 480 and the GeForce GTX 1060 in performance, with a 4% difference either way.

 

For Honor (3840 x 2160)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by For Honor‘s internal DirectX 12 benchmark.

At this ultra high resolution, the GeForce GTX 1060 finally pulled away, with a 7% performance advantage over the Radeon RX 580 (Amazon), which was 3% faster than the Radeon RX 480. Only the NVIDIA GeForce GTX 1070 was able to deliver an average frame rate above 30 fps.

Next Page > Mass Effect: Andromeda Benchmark Results, Our Verdict

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

Mass Effect: Andromeda (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in Mass Effect: Andromeda.

At this entry-level gaming resolution, the AMD Radeon RX 580 (Amazon) was no faster than the Radeon RX 480. They were both 9% slower than the GeForce GTX 1060 and 45% faster than the GeForce GTX 1050 Ti.

 

Mass Effect: Andromeda (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in Mass Effect: Andromeda.

With the jump in resolution, the Radeon RX 580 (Amazon) was just 1.5% faster than the Radeon RX 480. They were both 13% slower than the GeForce GTX 1060 and 48% faster than the GeForce GTX 1050 Ti.

 

Mass Effect: Andromeda (3840 x 2160)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in Mass Effect: Andromeda.

At this ultra high resolution, the Radeon RX 580 (Amazon) was now 2% faster than the Radeon RX 480. They were both 13% slower than the GeForce GTX 1060 and 53% faster than the GeForce GTX 1050 Ti. Only the NVIDIA GeForce GTX 1070 was able to deliver an average frame rate above 30 fps.

[adrotate group=”1″]

 

Our Verdict On The Radeon RX 580

It’s pretty obvious from our benchmark results that the AMD Radeon RX 580 (Amazon) is just a slightly faster version of its predecessor, the Radeon RX 480. That is no slight, as the Radeon RX 480 is a fast graphics card. In fact, it was so fast that NVIDIA was forced to introduce the GeForce GTX 1060 at a lower price point to compete.

The Radeon RX 580 (Amazon) reused the Polaris 10 GPU from the Radeon RX 480, running it and the GDDR5 memory at a 9% and 14% higher clock speeds respectively. However, it only delivered an performance boost of between 1.5% and 4%.

Obviously, you wouldn’t upgrade to the Radeon RX 580 (Amazon) if you already own the Radeon RX 480, or GeForce GTX 1060. But it would make for a great upgrade if you are using an older, slower graphics card. That is precisely why AMD is targeting the Radeon RX 500 Series at gamers with 3 years or older graphics cards.

Here is the key takeaway point from all these benchmarks. Like its predecessor, the Radeon RX 480, the AMD Radeon RX 580 (Amazon) is designed for 1440p gaming. With some tweaks to the graphics settings, you should have no problem achieving an average frame rate of 60 fps at that resolution.

[adrotate group=”2″]

Don’t forget to check out our other articles on AMD Radeon graphics technologies :

Go Back To > First Page | Computer Hardware + Systems | Home

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

The honor 8 Aurora Glass Smartphone Review

Boasting a unique Aurora Glass finish on its back, the new honor 8 smartphone is truly a work of art that has to be seen to be appreciated. But it is not just a good-looking smartphone. The honor 8 also features a dual-lens main camera, like its brother – the Huawei P9So join us as we take a close look at the honor 8 smartphone!

 

The honor 8 Smartphone Specifications

First up, we would like to point out that the honor 8 is available in 5 different models, with 3-4 colours, depending on the location. But they basically consist of two versions – a standard honor 8 model, and a premium honor 8 model. Check out their specifications :

Specificationshonor 8 (Standard)honor 8 (Premium)
ModelFRD-L02 (India), FRD-L04 (US), FRD-L09 (EU & Asia)FRD-L14 (US), FRD-L19 (EU & Asia)
Available ColoursPearl White, Midnight BlackSapphire Blue, Sunrise Gold
Display5.2" LTPS LCD display
- 1080 x 1920 resolution (423 ppi)
- 1500:1 contrast ratio, 96% NTSC colour gamut
5.2" LTPS LCD display
- 1080 x 1920 resolution (423 ppi)
- 1500:1 contrast ratio, 96% NTSC colour gamut
ProtectionNippon Electric Glass (NEG)Corning Gorilla Glass 3
Operating SystemAndroid 6.0 with Huawei Emotion UI 4.1Android 6.0 with Huawei Emotion UI 4.1
Processing PlatformHiSilicon Kirin 950HiSilicon Kirin 950
Processor4 x ARM Cortex-A72 (2.3 GHz)
4 x ARM Cortex-A53 (1.8 GHz)
i5 co-processor
4 x ARM Cortex-A72 (2.3 GHz)
4 x ARM Cortex-A53 (1.8 GHz)
i5 co-processor
Graphics ProcessorARM Mali-T880 (650 MHz)ARM Mali-T880 (650 MHz)
System Memory4 GB LPDDR4 memory4 GB LPDDR4 memory
Storage32 GB internal flash memory
microSD slot* (supports up to 128 GB)
* shared with SIM2 (L09)
64 GB internal flash memory
microSD slot* (supports up to 128 GB)
* shared with SIM2 (L19)
Selfie Camera8 MP sensor, with f/2.4 lens
- 1080p video recording
8 MP sensor, with f/2.4 lens
- 1080p video recording
Main CameraDual 12 MP cameras (monochrome + colour)
- Sony IMX286 sensor
- f/2.2 lens, dual LED flash and hybrid auto-focus
- 1080p video recording with EIS
Dual 12 MP cameras (monochrome + colour)
- Sony IMX286 sensor
- f/2.2 lens, dual LED flash and hybrid auto-focus
- 1080p video recording with EIS
GPSGPS / AGPS / GLONASSGPS / AGPS / GLONASS
SIM SupportSingle nano SIM (L02, L04)
Dual nano SIM, Dual Standby (L09)
Single nano SIM (L14)
Dual nano SIM, Dual Standby (L19)
Connectivity4G : LTE FDD LTE Cat6 : 51 Mbps (UL), 301 Mbps (DL)
3G : WCDMA
2 G : GSM
Bluetooth : 4.2 BLE
OTG, VoLTE
4G : LTE FDD LTE Cat6 : 51 Mbps (UL), 301 Mbps (DL)
3G : WCDMA
2 G : GSM
Bluetooth : 4.2 BLE
OTG, VoLTE
WiFi Capability802.11 a/b/g/n/ac (2.4 GHz, 5 GHz dual-band)
Hotspot supports up to 8 devices
802.11 a/b/g/n/ac (2.4 GHz, 5 GHz dual-band)
Hotspot supports up to 8 devices
USB PortUSB 2.0 with USB Type C interfaceUSB 2.0 with USB Type C interface
Fingerprint SensorYes, 3D fingerprint recognition, 0.4 second unlock with Smart Key customisationYes, 3D fingerprint recognition, 0.4 second unlock with Smart Key customisation
Infra-Red TransmitterYes, with Smart Control 4.0Yes, with Smart Control 4.0
NFC SupportYes, tag read/write, P2P transfer, mobile paymentYes, tag read/write, P2P transfer, mobile payment
AudioHiSilicon Hi6402 audio DSP with Dolby DTS and HiRes Audio support
Dual-microphone noise reduction
HiSilicon Hi6402 audio DSP with Dolby DTS and HiRes Audio support
Dual-microphone noise reduction
Battery2,900 mAh lithium-polymer battery
18 W fast-charging capability
2,900 mAh lithium-polymer battery
18 W fast-charging capability
Dimensions71.0 mm wide x 145.5 mm tall x 7.45 mm thick71.0 mm wide x 145.5 mm tall x 7.45 mm thick
Weight153 g153 g
Launch PriceRM 1,699 (~US$ 425)RM 1,899 (~US$ 475)

The standard honor 8 models are priced at RM 1,699 / ~US$ 425 with 32 GB of storage, and is available in either Pearl White or Midnight Black. The premium honor 8 models are priced at RM 1,899 / ~US$ 475 with 64 GB of storage, and either Sapphire Blue or Sunrise Gold colours.

 

Unboxing The honor 8 Smartphone

The honor 8 comes in a white two-piece cardboard box with a textured surface. The tall and thin design presumably allows more units to be placed side-by-side on a counter. The inner box slides downwards and is secured by a single clear security sticker. Be careful once that sticker is cut, because the inner box will slide right out!

[adrotate banner=”4″]

Inside the honor 8 box, you will find the following items :

  • One honor 8 smartphone
  • One set of honor 8 Quick Start Guide & Huawei warranty leaflet
  • One SIM card tray extractor pin
  • One set of six IMEI stickers
  • One 18 W Huawei quick travel charger
  • One USB 2.0 cable with USB Type C interface

Next Page > The honor 8 Up Close, Hybrid SIM, Fingerprint Sensor & OTG Support

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

The honor 8 Smartphone Up Close

The honor 8 has a large 5.2″ LTPS display that covers almost its entire front – with a 72% screen-to-body ratio. To achieve that, honor eschewed the use of physical buttons, in favour of virtual on-screen buttons. In fact, it only has two physical buttons – a Power button and a Volume rocker switch on the right side.

The honor 8 has an infra-red transmitter at the top, which can be used as a remote control in conjunction with the built-in Smart Controller app. It is also the first honor smartphone to feature the new Type C USB interface. It is still a USB 2.0 port, but the Type C interface is bidirectional – you no longer have to worry about plugging the cable in the wrong way.

Even though the honor 8 is the same size as the honor 5C, they managed to squeeze in a 2,900 mAh lithium-polymer battery while reducing its thickness to just 7.45 mm. They even fit in the 3.5mm audio port, albeit at the bottom, together with the USB port.

 

Hybrid SIM

The honor 8 has dual SIM, dual standby capability. However, only the FRD-L09 and FRD-L19 models come with the hybrid SIM slot that allows for a second nano SIM card. These two models are only sold in Europe and Asia. The other models only support a single nano SIM.

Of course, the hybrid SIM slot means you have to choose between using a second nano SIM card, or a microSD card. You can’t use both simultaneously.

[adrotate banner=”5″]

 

Fingerprint Sensor & OTG Support

The honor 8 boasts a new 3D fingerprint sensor that can actually detect the ridges of the finger. This allows for more accurate detection of your fingerprint. With an auto-learning ability, the honor 8’s fingerprint sensor can recognise your fingerprint and unlock the honor 8 in just 0.4 seconds.

On the other hand, the honor 8 (like the honor 5C) does not support USB OTG (On-The-Go) drives. If you plug any USB OTG flash drive into its USB port, it will not be detected. The honor 8 also requires you to use the Huawei HiSuite software to access its files from your computer. This also limits what you can copy out from the honor 8 smartphone.

Next Page > Cameras, Aurora Glass, Initial Setup, Memory & Storage

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

Cameras

The honor 8 has an 8 MP front camera with an f/2.4 lens. It will deliver a maximum photo resolution of 3264 x 2448 in the 4:3 format. Again, a wider 16:9 format will crop that to 3264 x 1840. It will record videos in 1080p quality with electronic image stabilisation.

Its killer feature is the dual-lens camera at the back. It is actually two individual cameras, each with a Sony IMX286 sensors with f/2.2 lens. One camera captures the image in monochrome, while the other captures the same image in colour. They are then combined to deliver sharper and more detailed pictures. The camera is supported by a dual LED flash, and a laser autofocus system.

The photos have a maximum resolution of 3968 x 2976 in the 4:3 format. If you prefer a wider format, the honor 8 will do that by cropping the image to 3968 x 2240 (16:9). It is capable of recording video in full HD quality (1080p) with electronic image stabilisation.

 

Aurora Glass

The honor 8 has a 2.5D glass body on both sides with diamond-cut aluminium alloy edges to create an ultra-narrow bezel. Its 3D grate design and 15-layer craftsmanship on the back of the device, combines with the glass/metal back to create what honor calls Aurora Glass.

The Aurora Glass back gives the honor 8 a very reflective surface at the back. The only catch – it naturally lacks the oleophobic coating of the front screen for a proper grip. However, that means its highly-reflective back is easily smudged by fingerprint smears, which detracts from its lustre.

 

The Initial Setup

This video will show you the initial setup process when you power up the honor 8 for the first time. If you do not need to restore a backup, the whole setup process should take about 5 minutes. Note how fast it takes to register a fingerprint.

[adrotate banner=”5″]

 

Available Memory & Storage

The honor 8 comes with 4 GB of LPDDR4 memory. The Android 6.0 operating system and honor-preloaded apps take up about 1.3 GB of memory, leaving you with 2.4 GB (60%) of free memory. Very good!

This is a premium honor 8 model (FRD-L19), which comes with 64 GB of internal flash memory. This is expandable via a microSD card (up to 128 GB). Fresh out of the box, the honor 8 has 53.2 GB of available space (excluding the system files), of which 51.38 GB is free to use.

In the standard honor 8 models which come with 32 GB of internal flash memory, we can estimate that they will have about 19 GB of free space.

Next Page > honor 8 Camera & Video Recording Performance

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

Camera Performance – Photos

The honor 8‘s dual-lens camera has not one, but two, 12 MP Sony IMX286 sensor with reasonably fast f/2.4 lens. The photos have a maximum resolution of 3968 x 2976 in the 4:3 format. Check out these photo samples.

The honor 8 delivered excellent resolution in these photos. If you blow them up to 100%, you can see the amount of detail that is not possible in other smartphone cameras. The edge resolution (between elements of different colour), in particular, was very impressive. The amount of noise is also very low, even in dark areas.

While the honor 8 does not have the most user-friendly camera app in the industry, the photos produced by its dual-lens camera is very, very impressive. Even without turning on HDR, the colour are very vivid and visually pop out.

[adrotate banner=”5″]

 

Camera Performance – Video Recording

The honor 8 can record full HD (1080p) ideos with electronic stabilisation. This is an improvement over the honor 5X, whose main camera does not support electronic stabilisation. Here is a sample video we recorded of a piano performance by Mr. Yang I-Deun.

And this is a sample video taken at a street market.

The honor 8 did a great job maintaining focus and controlling exposure in the videos we took. Even bright backgrounds were not overblown. The honor 8 does not support optical image stabilisation, probably because it would prevent the dual-lens system from being in sync, but its electronic image stabilisation still did a decent job.

You should also notice that it records better audio than most smartphones. The piano music sounds richer, even through the background noise. And in the street market footage, the ambient sounds sound clearer and more authentic.

Next Page > Work Performance, Battery Life & Recharging Performance

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

Work Performance – PCMark

We tested the honor 8‘s performance using PCMark, which simulates work applications like web browing, playing video, writing text and editing photos.

honor 8 ASUS ZenFone 3 Ultra HTC 10 Samsung Galaxy S7 edge
Work Performance Score
(Average)
6350 6313 6105 4892
Web Browsing 8073 6257 5782 6908
Video Playback 4396 4287 3321 3218
Writing 6231 7166 7463 4629
Photo Editing 7352 8262 9694 5563

The honor 8 is powered by the top-of-the-line Kirin 950 SoC. These PCMark results show that the Kirin 950 proved to be more than a match for the Snapdragon 820 that powers the HTC 10, and the Snapdragon 652 that powers the ASUS ZenFone 3 Ultra. In fact, it was 30% faster in this benchmark than Samsung Galaxy S7 edge‘s Exynos 8890!

 

Battery Performance – PCMark

We then tested the honor 8‘s battery performance using PCMark. It basically ran the PCMark work tests until its battery capacity dropped to 20%.

The honor 8 lasted just over 10 hours on the PCMark stress test. This is 1.5 hours less than the honor 5C, but is excellent considering it uses a more powerful processor and graphics processor, and comes with twice the memory. The HTC 10, for example, lasted just over 5.5 hours on the same test!

[adrotate banner=”5″]

 

Battery Recharging Performance

The honor 8 has a large 2,900 mAh lithium-polymer battery, with Fast Charging capability. honor claims that it will recharge from 0% to 47% in just 30 minutes. Let’s see if that’s true…

Our tests showed that the honor 8 will recharge from 17% to 100% in just 146 minutes (3 hours 26 minutes). This is decent, but not exceptional quick charging performance. It fully recharges the battery about 20% faster than the honor 5C, which does not support quick charging and uses only a 5 W charger.

That said, we can see that the honor 8 recharges fastest in the first 90%, before it tapers off to a slower recharging pace for the last 10 percent. Recharging 47% (from 17% to 64%) took 36 minutes in our test.

Next Page > Overall & Gaming Performance, Our Verdict & Award

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

Overall Performance – AnTuTu Benchmark

AnTuTu is a nice benchmark for testing various performance aspects of Android devices.

honor 8 ASUS ZenFone 3 Ultra HTC 10 Samsung Galaxy S7 edge
TOTAL SCORE 90737 76096 113813 131246
3D SCORE 17706 18816 50700 40824
– 3D (Marooned) 9829 10861 34778 23336
– 3D (Garden) 7877 7952 15922 17488
UX SCORE 34550 26794 30890 42765
– UX Data Secure 7338 5077 6319 8169
– UX Data Process 5513 4198 3379 5966
– UX Strategy Games 9745 7503 5433 10745
– UX Image Process 5017 6879 9521 13878
– UX I/O Performance 6937 3137 6238 4007
CPU SCORE 32255 23689 21915 37882
– CPU Mathematics 10483 6950 8494 11831
– CPU Common Use 8543 6311 5931 10875
– CPU Multi-Core 13229 10428 7490 15176
RAM 6226 6797 10308 9775

The honor 8 did well in AnTuTu, coming in only 20% and 31% slower than the HTC 10 and the Samsung Galaxy S7 edge respectively.

It was actually 12% and 47% faster than the HTC 10 in the UX and CPU tests, but its 65% slower 3D performance pulled down its overall results.

 

Gaming Performance – 3DMark

We tested the honor 8‘s gaming performance using 3DMark, using the Ice Storm Unlimited test.

honor 8 ASUS ZenFone 3 Ultra HTC 10 Samsung Galaxy S7 edge
Gaming Score 18659 17719 23171 28650
Graphics 19993 21124 27512 33685
Physics 15127 11328 14925 18810
Graphics Test 1 122.5 fps 115.2 fps 136.6 fps 162.9 fps
Graphics Test 2 67.4 fps 76.3 fps 106.4 fps 133.1 fps
Physics Test 48.0 fps 36.0 fps 47.4 fps 59.7 fps

The honor 8 is roughly equal to the ASUS ZenFone 3 Ultra in gaming performance. This means the ARM Mali-T880 GPU in the Kirin 950 is about as fast as the Qualcomm Adreno 510 GPU.

 

Our Verdict

The honor 8 is more than just a more powerful version of the honor 5C smartphone. Yes, they are both 5.2″ smartphones powered by Kirin processors from HiSilicon, a Huawei company. But the honor 8 is really a whole different beast targeting a different market.

The honor 8 is very fast, and comes with 4 GB of RAM and up to 64 GB of internal flash storage. On top of that, its dual-lens camera produces excellent photographs and videos. It even comes with a new 3D fingerprint sensor that activates the phone in just 0.4 seconds.

The honor 8 also looks the part of a premium smartphone, thanks to its Aurora Glass finish. This is honor’s top-of-the-line smartphone for those who want nothing but the best without paying through the nose.

Now, all that extra capability comes at a price, and it comes to no surprise that the honor 8 costs twice as much as the honor 5C. Even so, it costs less than half of what their competitors charge for equivalent smartphones.

[adrotate banner=”4″]

So are we impressed with the honor 8? Hell, yeah!

If we have any gripes, it is the lack of USB OTG support and the requirement to use the Huawei HiSuite to manage files using a computer. This unnecessarily makes it a pain to access the files in the honor 8 smartphone.

honor continues to impress us with their ability to focus on creating no-nonsense Android smartphones with practical features. Even the dual-lens camera proved to be a practical way to improve the quality of photos and videos taken using a smartphone.

We like the honor 8 enough to give it out coveted Editor’s Choice Award. Congratulations, honor!

 

Back To Start > honor 8 Specifications & Unboxing

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

The GeForce GTX 1050 & 1050 Ti Performance Comparison

On 20 October 2016, NVIDIA officially launched the new GeForce GTX 1050 and GeForce GTX 1050 Ti graphics cards. Built around the new NVIDIA GP107 GPU, the two new GeForce cards are designed for entry-level and eSports gaming at very attractive price points of US$99 and US$139 respectively. But just how fast are they? Find out for yourself in our GeForce GTX 1050 & GTX 1050 Ti performance comparison!

 

The Graphics Cards Compared

In this comparison, we will take a look at six different graphics cards launched in 2016 – 3 from AMD, and 3 from NVIDIA. We will pit the GeForce GTX 1050 against the Radeon RX 460, and the GeForce GTX 1050 Ti against the Radeon RX 470 (Review). We also included the higher-end Radeon RX 480 (Review) and GeForce GTX 1060 (Review) in this comparison.

SpecificationsRadeon RX 460GeForce GTX 1050GeForce GTX 1050 TiRadeon RX 470Radeon RX 480GeForce GTX 1060
Texture Units56404812814480
ROPs163232323248
Core Speed1090 MHz1354 MHz1290 MHz926 MHz1120 MHz1506 MHz
Boost Speed1200 MHz1455 MHz1392 MHz1206 MHz1266 MHz1708 MHz
Texture Fill Rate (Max)67.2 GT/s54.2 GT/s61.9 GT/s154.4 GT/s182.3 GT/s136.6 GT/s
Pixel Fill Rate (Max)19.2 GP/s46.6 GP/s44.5 GP/s38.6 GP/s40.5 GP/s82.0 GP/s
Memory Bus Width128-bits128-bits128-bits256-bits256-bits192-bits
Graphics Memory4 GB GDDR52 GB GDDR54 GB GDDR54 GB GDDR58 GB GDDR56 GB GDDR5
Memory Speed1750 MHz1752 MHz1752 MHz1650 MHz1750 MHz2000 MHz
Memory Bandwidth112 GB/s112 GB/s112 GB/s211 GB/s224 GB/s192 GB/s
TDP75 W75 W75 W120 W150 W120 W
Current PriceUS$ 99US$ 99US$ 139US$ 169US$ 239 (8 GB)US$ 249

 

Benchmarking Notes

Our graphics benchmarking test bed has the following specifications :

[adrotate banner=”4″]

Operating System : Microsoft Windows 10 64-bit

Processor : Intel Core i7 6700K processor running at 4.0 GHz

Motherboard : ASRock Z170 Extreme4

Memory : 8 GB DDR4-2133 memory (dual-channel)

Storage : 240 GB HyperX Savage SSD

Monitor : Dell P2415Q Ultra HD Monitor

We used the GeForce driver version 375.63 for the three NVIDIA graphics cards, and the Radeon Software 16.9.2 driver for the three AMD graphics cards used in our tests.

Okay, let’s get on with the GeForce GTX 1050 & 1050 Ti performance comparison!

Next Page > 3DMark DirectX 12 & Direct 11 Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


3DMark DirectX 12 Benchmark (2560 x 1440)

3DMark Time Spy is a new DirectX 12 benchmark that supports new API features like asynchronous compute, explicit multi-adapter, and multi-threading.

The GeForce GTX 1050 is virtually equivalent to the Radeon RX 460 in performance, with the GeForce GTX 1050 Ti delivering 23% better performance. The Radeon RX 470 was 60% faster than the GeForce GTX 1050 Ti, and twice as fast as the GeForce GTX 1050.

 

3DMark (1920 x 1080)

For Direct 11 performance, we started testing the graphics cards using 3DMark at the most common gaming resolution – 1920 x 1080.

The GeForce GTX 1050 and 1050 Ti did better in DirectX 11. In this test, they were 17.6% and 27.4% faster than the Radeon RX 460 respectively.

The Radeon RX 470, on the other hand, was now 48% faster than the GeForce GTX 1050 Ti, and 67% faster than the GeForce GTX 1050.

[adrotate banner=”5″]

 

3DMark (2560 x 1440)

Then we took 3DMark up a notch to the resolution of 2560 x 1440. Let’s take a look!

Surprisingly, the GeForce GTX 1050 and 1050 Ti did even better at this resolution. Despite having only 2 GB of GDDR5 memory, the GeForce GTX 1050 was 24% faster than the Radeon RX 460. The GeForce GTX 1050 Ti was now 34% faster than the Radeon RX 460.

They also edged a little closer in performance to the Radeon RX 470, which was now 47% faster than the GeForce GTX 1050 Ti, and 60% faster than the GeForce GTX 1050.

 

3DMark (3840 x 2160)

This is torture, even for the GeForce GTX 980 Ti and the GeForce GTX 1070!

It is definitely a very, very bad idea to play any game at this resolution on the GeForce GTX 1050 and 1050 Ti. Especially the GeForce GTX 1050, which only has 2 GB of memory. In this resolution, the Radeon RX 460 (which has 4 GB of memory) was actually 7% faster! The GeForce GTX 1050 Ti was now 33% faster than the Radeon RX 460.

The Radeon RX 470 was now 49% faster than the GeForce GTX 1050 Ti, and twice as fast as the GeForce GTX 1050.

Next Page > Ashes of the Singularity & Warhammer Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


Ashes of the Singularity (1920 x 1080)

We tested Ashes of the Singularity in the DirectX 12 mode, which supports the new Asynchronous Compute feature. We started with the full HD resolution.

The GeForce GTX 1050 and 1050 Ti were 6% and 17% faster than the Radeon RX 460 respectively. All three delivered playable frame rates of over 30 fps at this resolution.

 

Ashes of the Singularity (2560 x 1440)

We then took Ashes of the Singularity up a notch to the resolution of 2560 x 1440. Let’s see how the cards fare…

Only the GeForce GTX 1050 Ti was fast enough to deliver a playable frame rate above 30 fps at this resolution. The Radeon RX 460 was actually slightly (2%) faster than the GeForce GTX 1050, probably due to its larger memory size.

[adrotate banner=”5″]

 

Warhammer (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by Total War : Warhammer‘s internal DirectX 12 benchmark.

The GeForce GTX 1050 and 1050 Ti were 2% and 15% faster than the Radeon RX 460 respectively. All three delivered playable frame rates of about 40 fps at this resolution.

 

Warhammer (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by Total War : Warhammer‘s internal DirectX 12 benchmark.

The GeForce GTX 1050 and 1050 Ti did better at this resolution, coming in 9% and 20% faster than the Radeon RX 460 respectively. Only the GeForce GTX 1050 Ti delivered (barely) playable frame rates at this resolution.

Next Page > The Witcher 3 & Fallout 4 Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


The Witcher 3 (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in The Witcher 3.

The GeForce GTX 1050 and GeForce GTX 1050 Ti did well in this game, coming in 17% and 32% faster than the Radeon RX 460 respectively. The two new GeForce cards delivered reasonably playable frame rates.

 

The Witcher 3 (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in The Witcher 3.

The GeForce GTX 1050 and GeForce GTX 1050 Ti maintained their performance advantage over the Radeon RX 460, but their frame rates were poor. Definitely not a good idea to play The Witcher 3 at 1440p with these cards.

[adrotate banner=”5″]

 

Fallout 4 (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in Fallout 4.

The GeForce GTX 1050 and GeForce GTX 1050 Ti performed very well in Fallout 4, both delivering 20% better performance than the Radeon RX 460. This is the perfect resolution for all three cards.

 

Fallout 4 (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in Fallout 4.

Surprisingly, the GeForce GTX 1050 and GeForce GTX 1050 Ti extended their lead over the Radeon RX 460, with 22% and 24% better performance respectively. Both new GeForce cards actually delivered playable frame rates at this resolution.

Next Page > Our Verdict, Lowest Prices

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


Our Verdict

Hitherto, AMD had a real advantage in the entry-level and eSports gaming market with their Radeon RX 460 and Radeon RX 470 graphics cards. Priced at US$109 and US$179 respectively (at launch), they were much more affordable options for gaming at 1080p or lower resolutions.

That changed with the launch of the GeForce GTX 1050 and GeForce GTX 1050 Ti graphics cards.

The GeForce GTX 1050 was targeted squarely at the Radeon RX 460, with the GeForce GTX 1050 Ti offering slightly better performance and more graphics memory for those who can afford an extra US$40.

AMD is cognisant of the danger the new GeForce cards pose. So they preemptively slashed the prices of their Radeon RX 460 and Radeon RX 470 cards to make them more competitive. Did they succeed?

GeForce GTX 1050

Both the GeForce GTX 1050 and the Radeon RX 460 are priced at US$99. However, the GeForce GTX 1050 either matches the Radeon RX 460 in performance, or outperforms it by up to 22%. It does this despite having only 2 GB of memory.

[adrotate banner=”4″]

So if you are looking for a graphics card at the US$99 price point, the obvious choice is the GeForce GTX 1050.

GeForce GTX 1050 Ti

The GeForce GTX 1050 Ti is roughly 10-15% faster than the GeForce GTX 1050, and has twice as much graphics memory. However, that is a limited advantage since it is still best used for 1080p gaming.

If you can cough up the extra US$40, you might as well cough up an extra US$30 to get the newly-repriced Radeon RX 470. That will buy you 45-70% better performance, enough for 1440p gaming.

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!


The XFX Radeon RX 470 RS Black Edition Review

AMD launched their Polaris GPU architecture to great fanfare with the introduction of the Radeon RX 480. It offered an affordable 1440p gaming alternative to the NVIDIA GeForce GTX 1070. The AMD Radeon RX 470 was released later, offering 1080p gaming performance at even lower cost and power consumption.

Today, we are going to take a look at the XFX Radeon RX 470 RS Black Edition (RX-470P4LDB6) graphics card.  This is no ordinary Radeon RX 470 graphics card. It is factory-overclocked and boasts a really unique cooler. Check it out!

 

The XFX Radeon RX 470 RS Black Edition Specification Comparison

How does the XFX Radeon RX 470 RS Black Edition compare against a standard AMD Radeon RX 470 graphics card? Take a look!

SpecificationsStandard AMD Radeon RX 470XFX Radeon RX 470 RS Black EditionDifference
Stream Processors20482048-
Texture Units128128-
ROPs3232-
GPU Clock Speed926 MHz base clock
1206 MHz boost clock
1256 MHz+ 4% to +35.6%
Texture Fill Rate118 to 154 GT/s160 GT/s+ 4% to +35.6%
Pixel Fill Rate30 to 39 GP/s40 GP/s+ 4% to +35.6%
Memory Bus Width256-bits256-bits-
Graphics Memory4 GB GDDR54 GB GDDR5-
Memory Clock Speed1650 MHz1750 MHz+ 6%
Memory Bandwidth211 GB/s224 GB/s+ 6%
TDP120 WNot stated (between 120 - 150 W)-
Cooling SolutionDual slot cooler with single fanXFX Ghost 4.0 coolerBetter cooling

For more specifications, please take a look at our Desktop Graphics Card Comparison Guide.

[adrotate banner=”5″]

 

Unboxing The XFX Radeon RX 470 RS Black Edition

Here is our unboxing video of the XFX Radeon RX 470 RS Black Edition graphics card. This is exactly what you can expect if you purchase one.

The XFX Radeon RX 470 RS Black Edition graphics card comes in an overly large cardboard box. Inside, you will only find the XFX Radeon RX 470 RS Black Edition graphics card, a combined installation guide and warranty card leaflet, a driver CD and a 4-pin to 6-pin power adaptor.

 

The XFX Radeon RX 470 RS Black Edition Overview

In this video, we are going to take a quick look at the XFX Radeon RX 470 RS Black Edition graphics card, and its unique Ghost 4.0 cooler.

Next Page > XFX Radeon RX 470 RS Black Edition Up Close, Ghost 4.0 Cooler

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participate in the Tech ARP Forums, or even donate to our fund. Any help you can render is greatly appreciated!

The XFX Radeon RX 470 RS Black Edition Up Close

The XFX Radeon RX 470 RS Black Edition is a large card – the same size as the AMD Radeon RX 480. The shroud of the XFX Ghost 4.0 cooler extends a little over the card itself, to accommodate the two large 85mm fans. The back is protected and stiffened by a solid aluminium backplate, which also serves as a secondary heatsink.

The XFX logo is etched into the backplate, with a white XFX logo on the top of the shroud. However, case modders should note that it does not have any LED lighting – the XFX logo will not light up when the card is powered up. The third XFX logo can be found as part of the exhaust vent, next to the DVI port.

[adrotate banner=”5″]

 

The XFX Ghost 4.0 Cooler

The XFX Ghost 4.0 cooler is, arguably, the highlight of this model. It features a unibody heatsink, with composite heatpipes. The unibody design improves its ability to cool the VRM and GDDR5 memory modules by 40%, while reducing fan noise by 5%.

The composite heatpipe design, on the other hand, combines the liquid and capillary action heatpipe in a single pipe. This increases the total heatpipe surface area by 30%, which greatly improves its ability to quickly transfer heat.

The other cool thing about the XFX Ghost 4.0 cooler are its two 85 mm fans, pun intended. With a power output of 4.2 W, two of these FirstD FDC10H12S9-C fans can push a considerable amount of air through the heatsink. But what’s really cool are its patented hard-swap design, and its load-sensing technology.

The hard swap design allows for easy detachment and installation of the two fans. They are locked in place by just two clips. You can swap them out for more powerful fans, or even fans with LED lights. It also allows you to easily remove them to clean the blades or even the heatsink fins underneath. Very useful!

The other cool feature is the fans’ auto load-sensing capability. They will automatically adjust their speed according to the workload. During light loads (watching videos, using office applications), the XFX Radeon RX 470 RS Black Edition runs cool enough that both fans do not spin at all, giving you absolute silence. But when you start gaming, the fans start up and increase their speed to meet the demand automatically.

Next Page > Benchmarking Notes, Cooling Performance, Noise Level

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participate in the Tech ARP Forums, or even donate to our fund. Any help you can render is greatly appreciated!

Benchmarking Notes

Our graphics benchmarking test bed has the following specifications :

Operating System : Microsoft Windows 10 64-bit

Processor : Intel Core i7 6700K processor running at 4.0 GHz

Motherboard : ASRock Z170 Extreme4

Memory : 8 GB DDR4-2133 memory (dual-channel)

Storage : 240 GB HyperX Savage SSD

Monitor : Dell P2415Q Ultra HD Monitor

We used the GeForce driver version 372.54 for all three NVIDIA graphics cards, and the Radeon Software 16.8.2 driver for the three AMD graphics cards used in our tests.

 

The XFX Radeon RX 470 RS Black Edition Cooling Performance

The AMD Radeon RX 470 uses the AMD Polaris 10 GPU, which is fabricated on the 14 nm FinFET process. The smaller process technology allows AMD to run the Radeon RX 470 GPU at 1.2 GHz with a TDP of just 120 W. This allows for a much cooler card, and hopefully, better overclocking potential.

With its Ghost 4.0 cooler, the XFX Radeon RX 470 RS Black Edition should deliver significantly better cooling performance than the stock Radeon RX 470 card. While we don’t have one at hand, we do have a number of other AMD and NVIDIA cards. Take a look at their peak exhaust temperature (as measured at their exhaust vents).

Note that these are not the recorded temperatures, but how much hotter the exhaust air is above ambient temperature.

The XFX Radeon RX 470 RS Black Edition is a surprisingly cool-running card, producing much cooler exhaust air than the Radeon RX 480, or the GeForce GTX 1060. There is a reason for this though, which we will examine in the next section…

[adrotate banner=”5″]

 

The XFX Radeon RX 470 RS Black Edition Noise Level

We recorded the XFX Radeon RX 470 RS Black Edition as it was running the 3DMark Fire Strike Ultra benchmark for the third time.

As you can hear, the two fans do produce a considerable amount of noise. That is the secret (and downside) to its significantly better cooling performance. Of course, this is an open testbed, so the noise will be muted in an enclosed chassis.

Next Page > 3DMark DirectX 12 & Direct 11 Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participate in the Tech ARP Forums, or even donate to our fund. Any help you can render is greatly appreciated!

3DMark DirectX 12 Benchmark (2560 x 1440)

3DMark Time Spy is a new DirectX 12 benchmark that supports new API features like asynchronous compute, explicit multi-adapter, and multi-threading.

The XFX Radeon RX 470 RS Black Edition is about 35% faster than the Radeon R9 380 – amazing performance for a card that costs just US$ 219. In fact, it is just 8% slower than the AMD Radeon RX 480 (8 GB), and 12% slower than the NVIDIA GeForce GTX 1060!

 

3DMark (1920 x 1080)

For Direct 11 performance, we started testing the graphics cards using 3DMark at the most common gaming resolution – 1920 x 1080.

The XFX Radeon RX 470 RS Black Edition was 40% faster than the Radeon R9 380. That puts it just 7% slower than the Radeon RX 480 (8 GB).

[adrotate banner=”5″]

 

3DMark (2560 x 1440)

Then we took 3DMark up a notch to the resolution of 2560 x 1440. Let’s take a look!

At this resolution, the extra memory of the Radeon RX 480 (8 GB) starts to pay off. It was now 13% faster than the XFX Radeon RX 470 RS Black Edition, which only has 4 GB of GDDR5 memory.

 

3DMark (3840 x 2160)

This is torture, even for the GeForce GTX 980 Ti and the GeForce GTX 1070!

The XFX Radeon RX 470 RS Black Edition was now 16% slower than the Radeon RX 480 (8 GB). Not that it matters. The frame rates are completely unplayable for all cards at this resolution.

Next Page > Ashes of the Singularity & Warhammer Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participate in the Tech ARP Forums, or even donate to our fund. Any help you can render is greatly appreciated!

Ashes of the Singularity (1920 x 1080)

We tested Ashes of the Singularity in the DirectX 12 mode, which supports the new Asynchronous Compute feature. We started with the full HD resolution.

The XFX Radeon RX 470 RS Black Edition did very well in this test, virtually tying with the Radeon RX 480 (8 GB) and GeForce GTX 1060!

 

Ashes of the Singularity (2560 x 1440)

We then took Ashes of the Singularity up a notch to the resolution of 2560 x 1440. Let’s see how the cards fare…

Again, the XFX Radeon RX 470 RS Black Edition did very well. It tied the Radeon RX 480 (8 GB) and GeForce GTX 1060.

 

Ashes of the Singularity (3840 x 2160)

Finally, let’s see how the cards perform with Ashes of the Singularity running at the Ultra HD resolution of 3840 x 2160.

Surprisingly, the XFX Radeon RX 470 RS Black Edition continued to match the Radeon RX 480 (8 GB) and the GeForce GTX 1060 in performance. It was actually able to deliver playable frame rates even at such a high resolution.

[adrotate banner=”5″]

 

Warhammer (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by Total War : Warhammer‘s internal DirectX 12 benchmark.

The XFX Radeon RX 470 RS Black Edition delivered frame rates in excess of 60 fps. It actually tied the GeForce GTX 1060 in performance, and came in just 7% slower than the Radeon RX 480 (8 GB).

 

Warhammer (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by Total War : Warhammer‘s internal DirectX 12 benchmark.

At this higher resolution, the XFX Radeon RX 470 RS Black Edition slipped under 50 fps. Still pretty decent performance. It was now 4% slower than the GeForce GTX 1060, and 8% slower than the Radeon RX 480 (8 GB).

 

Warhammer (3840 x 2160)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by Total War : Warhammer‘s internal DirectX 12 benchmark.

The Ultra HD resolution proved too much for most cards. Only the GeForce GTX 1070 and the GeForce GTX 980 Ti delivered playable frame rates. The XFX Radeon RX 470 RS Black Edition was now 9% slower than the Radeon RX 480 (8 GB), and 11% slower than the GeForce GTX 1060.

Next Page > The Witcher 3 & Fallout 4 Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participate in the Tech ARP Forums, or even donate to our fund. Any help you can render is greatly appreciated!

The Witcher 3 (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in The Witcher 3.

The XFX Radeon RX 470 RS Black Edition performed well at this resolution, delivering an average frame rate of 60 fps. That made it 9% slower than the Radeon RX 480 (8 GB), and 14% slower than the GeForce GTX 1060.

 

The Witcher 3 (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in The Witcher 3.

When the resolution increased to 1440p, all cards took a massive hit in frame rate. Only the GeForce GTX 1070 and the GeForce GTX 980 Ti delivered average frame rates above 60 fps.

The XFX Radeon RX 470 RS Black Edition had a playable average frame rate of 41 fps. It remained 9% slower than the Radeon RX 480 (8 GB), and 14% slower than the GeForce GTX 1060.

 

The Witcher 3 (3840 x 2160)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in The Witcher 3.

The Ultra HD resolution is a real torture, even humbling the GeForce GTX 1070 and the GeForce GTX 980 Ti.

The XFX Radeon RX 470 RS Black Edition remained 9% slower than the Radeon RX 480 (8 GB), but was now 16% slower than the GeForce GTX 1060.

[adrotate banner=”5″]

 

Fallout 4 (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in Fallout 4.

Fallout 4 is a relatively “easy” for most graphics cards at the resolution of 1920 x 1080. The XFX Radeon RX 470 RS Black Edition had no problem delivering an average frame rate of 78 fps. That makes it just 4% slower than the Radeon RX 480 (8 GB), but 17% slower than the GeForce GTX 1060.

 

Fallout 4 (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in Fallout 4.

When we took the resolution to the next level though, it really cut down on frame rates. The XFX Radeon RX 470 RS Black Edition‘s average frame rate of 56 fps still very good, being 27% faster than the Radeon R9 380. However, it was now 8% slower than the Radeon RX 480 (8 GB), but 19% slower than the GeForce GTX 1060.

 

Fallout 4 (3840 x 2160)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in Fallout 4.

The 4K resolution really taxed the cards. Even the GeForce GTX 1070 could not deliver an average frame rate of 60 fps.

The XFX Radeon RX 470 RS Black Edition only managed an average frame rate of 31 fps. That made it 7% slower than the Radeon RX 480 (8 GB), but 18% slower than the GeForce GTX 1060.

Next Page > Our Verdict, Specifications & Lowest Price

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participate in the Tech ARP Forums, or even donate to our fund. Any help you can render is greatly appreciated!

Our Verdict

The AMD Radeon RX 470 was always meant for 1080p gaming, but XFX managed to take it one step further with the XFX Radeon RX 470 RS Black Edition. Thanks to the factory-overclocking of the GPU and the GDDR5 memory, it offers performance close to that of the AMD Radeon RX 480.

In our real world benchmarks, the XFX Radeon RX 470 RS Black Edition was proven to be capable of delivering good frame rates even at 2560 x 1440. 1440p gaming is now possible with the Radeon RX 470!

Of course, the extra performance comes at a cost. The XFX Radeon RX 470 RS Black Edition has a launch price of US$219. That’s a US$40 premium over a standard Radeon RX 470, and a US$20 premium over a 4GB Radeon RX 480.

What you get in return is a well-made, factory-overclocked Radeon RX 470 card, with a powerful load-sensing cooler and a solid aluminium backplate. The hard-swappable fans are also a nice touch, allowing for very easy upgrades, replacement and maintenance.

Of course, you may just decide to skip these niceties and top up another US$10 for an 8GB Radeon RX 480. That will buy you a slightly faster card, with more overclocking potential.

SpecificationsStandard AMD Radeon RX 470XFX Radeon RX 470 RS Black EditionDifference
Stream Processors20482048-
Texture Units128128-
ROPs3232-
GPU Clock Speed926 MHz base clock
1206 MHz boost clock
1256 MHz+ 4% to +35.6%
Texture Fill Rate118 to 154 GT/s160 GT/s+ 4% to +35.6%
Pixel Fill Rate30 to 39 GP/s40 GP/s+ 4% to +35.6%
Memory Bus Width256-bits256-bits-
Graphics Memory4 GB GDDR54 GB GDDR5-
Memory Clock Speed1650 MHz1750 MHz+ 6%
Memory Bandwidth211 GB/s224 GB/s+ 6%
TDP120 WNot stated (between 120 - 150 W)-
Cooling SolutionDual slot cooler with single fanXFX Ghost 4.0 coolerBetter cooling

For more specifications, please take a look at our Desktop Graphics Card Comparison Guide.

[adrotate banner=”5″]

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participate in the Tech ARP Forums, or even donate to our fund. Any help you can render is greatly appreciated!

AMD Radeon RX 480 CrossFire Performance Comparison

You have seen our Radeon RX 480 review, so today, we are going to take a look at its CrossFire performance. For the uninitiated, that’s two Radeon RX 480 graphics cards running together.

You may recall that when Chief Architect of the Radeon Technologies Group, Raja Koduri, first revealed the Radeon RX 480, he made a startling claim that two Radeon RX 480 cards only utilized 51% of their processing capabilities to beat the GeForce GTX 1080 graphics card.

That was later clarified by Robert Hallock as a CPU-limited result. He also revealed that two Radeon RX 480 cards will deliver 83% better performance than a single Radeon RX 480 card.

Today, we are going to check out those claims in a variety of benchmarks and games. Let’s get down to it!

 

The Graphics Cards

Specifications Radeon R9 380 Radeon RX 480 GeForce GTX 1060 GeForce GTX 980 Ti GeForce GTX 1070 Radeon RX 480 CrossFire
Textures Per Clock 112 144 80 176 120 288
Pixels Per Clock 32 32 48 96 64 64
Clock Speed 970 MHz 1120 – 1266 MHz 1506 – 1708 MHz 1000 – 1075 MHz 1506 – 1683 MHz 1120 – 1266 MHz
Texture Fill Rate 108.6 GT/s 161.3 – 182.3 GT/s 120.5 – 136.6 GT/s 176.0 – 189.2 GT/s 180.7 – 202.0 GT/s 322.6 – 364.6 GT/s
Pixel Fill Rate 31.0 GP/s 35.8 – 40.5 GP/s 72.3 – 82.0 GP/s 96.0 – 104.5 GP/s 96.4 – 107.7 GP/s 71.7 – 81.0 GP/s
Graphics Memory 4 GB GDDR5 8 GB GDDR5 6 GB GDDR5 6 GB GDDR5 8 GB GDDR5 16 GB GDDR5

(8 GB effective)

Memory Bus Width 256-bits 256-bits 192-bits 384-bits 256-bits 256-bits x 2
Memory Speed 1425 MHz 1750 MHz 2000 MHz 1752.5 MHz 2000 MHz 1750 MHz
Memory Bandwidth 182.4 GB/s 224.0 GB/s 192.0 GB/s 336.5 GB/s 256.0 GB/s 448.0 GB/s
TDP 190 W 150 W 120 W 250 W 150 W 300 W
[adrotate banner=”5″]

 

Benchmarking Notes

Our graphics benchmarking test bed has the following specifications :

Operating SystemMicrosoft Windows 10 64-bit

ProcessorIntel Core i7 6700K processor running at 4.0 GHz

MotherboardASRock Z170 Extreme4

Memory8 GB DDR4-2133 memory (dual-channel)

Storage240 GB HyperX Savage SSD

MonitorDell P2415Q Ultra HD Monitor

We used the GeForce driver version 372.54 for all three NVIDIA graphics cards used in our tests.

We used the Radeon Software 16.8.2 driver for the AMD graphics cards used in our tests. In addition, we enabled the Compatibility Mode for the Radeon RX 480 cards to ensure that they keep to their rated 150 W TDP.

Next Page > 3DMark DirectX 12 and DirectX 11 Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participate in the Tech ARP Forums, or even donate to our fund. Any help you can render is greatly appreciated!

3DMark DirectX 12 Benchmark (2560 x 1440)

3DMark Time Spy is a new DirectX 12 benchmark that was released just days ago, just in time for this review. It supports new API features like asynchronous compute, explicit multi-adapter, and multi-threading.

In this DirectX 12 benchmark, the Radeon RX 480 CrossFire was 93% faster than the single Radeon RX 480. Very impressive. This allows the Radeon RX 480 CrossFire to beat the GeForce GTX 1070 by 33%. This gives the Radeon RX 480 CrossFire a slight (5.7%) price-performance advantage over the GeForce GTX 1070, albeit at twice the power consumption.

 

3DMark (1920 x 1080)

For Direct 11 performance, we started testing the graphics cards using 3DMark at the most common gaming resolution – 1920 x 1080.

At this relatively CPU-limited test, the Radeon RX 480 CrossFire was 95% faster than the single Radeon RX 480. It was also 31%-35% faster than the GeForce GTX 1070, 79-86% faster than the GeForce GTX 1060.

Note that the Radeon RX 480 CrossFire ended up about 12% slower than both the GeForce GTX 1070 and the GeForce GTX 980 Ti in the Combined Test, probably because it was CPU-limited.

[adrotate banner=”5″]

 

3DMark (2560 x 1440)

Then we took 3DMark up a notch to the resolution of 2560 x 1440. Let’s take a look!

As the higher 1440p resolution, the Radeon RX 480 CrossFire was 96% faster than the single Radeon RX 480. However, its performance advantage over the GeForce GTX 1070 and GeForce GTX 1060 dropped to 27% and 74% respectively.

At this point, the Radeon RX 480 CrossFire loses its price-performance advantage over the GeForce GTX 1070. However, it still maintains a comfortable price-performance advantage over the GeForce GTX 1080.

 

3DMark (3840 x 2160)

This is a torture test, perfect for comparing the GeForce GTX 1070 and the Radeon RX 480 CrossFire.

At the 4K resolution, the Radeon RX 480 CrossFire was 23% faster than the GeForce GTX 1070, and 73% faster than the GeForce GTX 1060. If we had a GeForce GTX 1080, the Radeon RX 480 CrossFire would likely be slightly faster, with a significant price-performance advantage, albeit with significantly higher power consumption.

Next Page > Ashes of the Singularity, Total War: Warhammer Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

Ashes of the Singularity FAILED!

We tested Ashes of the Singularity in the DirectX 12 mode, which not only supports the new Asynchronous Compute feature, but is necessary to support the CrossFire mode for the two Radeon RX 480 cards.

Unfortunately, Ashes of the Singularity kept crashing whenever multi-GPU mode was enabled. There was simply no way to get it to run reliably, even though we tried 3 different driver versions.

As our GeForce GTX 1070 review shows, we have no issues running it with single graphics cards from both AMD and NVIDIA. It only failed when multi-GPU support was enabled. We will update this section when we finally get CrossFire mode running on Ashes of the Singularity.

 

Warhammer (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by Total War : Warhammer‘s internal DirectX 12 benchmark.

Looks like CrossFire isn’t working in Total War : Warhammer. The Radeon RX 480 CrossFire was actually 6.5% slower than the single Radeon RX 480 graphics card. It was just slightly faster than the GeForce GTX 1060.

 

Warhammer (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by Total War : Warhammer‘s internal DirectX 12 benchmark.

At the higher 1440p resolution, the Radeon RX 480 CrossFire was 7.7% slower than the single Radeon RX 480 graphics card, and 3.7% slower than the GeForce GTX 1060.

 

Warhammer (3840 x 2160)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, recorded by Total War : Warhammer‘s internal DirectX 12 benchmark.

At the 4K resolution, the Radeon RX 480 CrossFire was 4.9% slower than the single Radeon RX 480 graphics card, and 6.9% slower than the GeForce GTX 1060. It would definitely be a good idea to disable CrossFire mode when you play Total War : Warhammer.

Next Page > The Witcher 3 and Fallout 4 Benchmark Results

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

The Witcher 3 (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in The Witcher 3.

The CrossFire mode worked in The Witcher 3 though. The Radeon RX 480 CrossFire achieved an average frame rate in excess of 100 fps. That makes its 59% faster than the single Radeon RX 480 graphics card, 49.6% faster than the GeForce GTX 1060, and 5% faster than the GeForce GTX 1070.

 

The Witcher 3 (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in The Witcher 3.

The Radeon RX 480 CrossFire maintained its performance lead at the 1440p resolution. It was 59% faster than the single Radeon RX 480 graphics card, 47.8% faster than the GeForce GTX 1060, and 5.6% faster than the GeForce GTX 1070.

 

The Witcher 3 (3840 x 2160)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in The Witcher 3.

The Radeon RX 480 CrossFire increased its performance lead at the 4K resolution. It was now 64.4% faster than the single Radeon RX 480 graphics card, 52.3% faster than the GeForce GTX 1060, and 8.9% faster than the GeForce GTX 1070.

[adrotate banner=”5″]

 

Fallout 4 (1920 x 1080)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in Fallout 4.

CrossFire did not work in Fallout 4, at least not in 1080p. The Radeon RX 480 CrossFire was actually 2.7% slower than the single Radeon RX 480 graphics card, and 16% slower than the GeForce GTX 1060.

 

Fallout 4 (2560 x 1440)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in Fallout 4.

At the higher 1440p resolution, the CrossFire mode finally kicked in. The Radeon RX 480 CrossFire actually became 24% faster than the single Radeon RX 480 graphics card, and 9.6% faster than the GeForce GTX 1060.

 

Fallout 4 (3840 x 2160)

This chart shows you the minimum and maximum frame rates, as well as the average frame rate, that FRAPS recorded in Fallout 4.

Amazingly, when we hit 4K, the CrossFire mode really showed its mettle. The Radeon RX 480 CrossFire was now 62% faster than the single Radeon RX 480 graphics card, and 43.2% faster than the GeForce GTX 1060.

The Radeon RX 480 CrossFire even edged out the GeForce GTX 1070, and beat the GeForce GTX 980 Ti by 8.7%.

Next Page > Radeon RX 480 CrossFire Performance Summary, Our Opinion

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

Performance Summary

Here is a summary of our benchmark results. We highlighted the benchmarks in which the CrossFire mode worked (in green), and when it didn’t work (in red).

Benchmarks Radeon R9 380 Radeon RX 480 GeForce GTX 1060 GeForce GTX 980 Ti GeForce GTX 1070 Radeon RX 480 CrossFire
Time Spy (1440p) Slower by 65.1% Slower by 48.2% Slower by 45.8% Slower by 36.5% Slower by 24.8% Baseline
Fire Strike (1080p) Slower by 64.0% Slower by 48.8% Slower by 45.2% Slower by 28.7% Slower by 25.1% Baseline
Fire Strike Extreme (1440p) Slower by 63.4% Slower by 49.0% Slower by 42.4% Slower by 25.1% Slower by 21.2% Baseline
Fire Strike Ultra (2160p) Slower by 61.2% Slower by 48.3% Slower by 42.1% Slower by 22.5% Slower by 18.8% Baseline
Ashes of the Singularity Failed
Total War: Warhammer (1080p) Slower by 26.1% Faster by 7.0% Slower by 0.5% Faster by 30.7% Faster by 36.1% Baseline
Total War: Warhammer (1440p) Slower by 26.4% Faster by 8.4% Faster by 3.9% Faster by 40.3% Faster by 43.8% Baseline
Total War: Warhammer (2160p) Slower by 29.4% Faster by 5.1% Faster by 7.5% Faster by 48.6% Faster by 51.0% Baseline
The Witcher 3 (1080p) Slower by 57.6% Slower by 37.2% Slower by 33.1% Slower by 8.0% Slower by 4.8% Baseline
The Witcher 3 (1440p) Slower by 57.0% Slower by 37.2% Slower by 33.1% Slower by 7.8% Slower by 5.3% Baseline
The Witcher 3 (2160p) Slower by 57.5% Slower by 39.2% Slower by 34.4% Slower by 10.2% Slower by 8.1% Baseline
Fallout 4 (1080p) Slower by 20.5% Faster by 2.8% Faster by 19.1% Faster by 27.8% Faster by 29.6% Baseline
Fallout 4 (1440p) Slower by 41.7% Slower by 19.6% Slower by 8.7% Faster by 16.7% Faster by 22.6% Baseline
Fallout 4 (2160p) Slower by 54.3% Slower by 38.1% Slower by 30.2% Slower by 8.0% Slower by 1.1% Baseline

 

Our Opinion

The Radeon RX 480 CrossFire showed great promise in the 3DMark benchmarks. It was able to deliver 93% to 96% better performance than a single Radeon RX 480 graphics card.

The actual boost in frame rate was smaller, of course, due to CPU limits. But it proved to be faster than the GeForce GTX 1070 by 15% to 18% in DirectX 11, and 26% in DirectX 12. Very impressive.

Unfortunately, we do not have an NVIDIA GeForce GTX 1080 in our benchmark suite, but we know that it is roughly 20%-25% faster than the GeForce GTX 1070. So we can guesstimate that the Radeon RX 480 CrossFire will be slightly faster than the GeForce GTX 1080 in DirectX 12, and slightly slower in DirectX 11.

[adrotate banner=”4″]

From a price-performance perspective, the Radeon RX 480 CrossFire ties with the GeForce GTX 1070, if we only take into account the 3DMark results. It loses out to the GeForce GTX 1070 when it comes to actual games, particularly in games that don’t work well or at all with CrossFire.

As you can tell from the table above, the reliability of the CrossFire mode is still quite iffy. That is really too bad, because the 3DMark results show that the Radeon RX 480 CrossFire has great potential, particularly in DirectX 12.

If AMD can get the CrossFire mode to work in all games, the Radeon RX 480 CrossFire is a great alternative to the GeForce GTX 1080. It offers equivalent performance at a 20% discount (US$478 vs. US$599), albeit with much higher power consumption (300W vs. 180W).

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participating in the Tech ARP Forums, or even donating to our fund. Any help you can render is greatly appreciated!

Odd Radeon RX 480 AOTS Benchmark Results Explained

When we posted about the revelation of the AMD Radeon RX 480, RX 470 and RX 460 graphics cards, we pointed out the odd Radeon RX 480 AOTS benchmark results that were presented by Raja Koduri, Senior Vice President and Chief Architect, Radeon Technologies Group.

The good news is we now understand how AMD derived those results. The bad news is it casts AMD in a rather bad light, and detracts from the great value proposition that the Radeon RX 480 brings to the table. But first, a quick recap of the Radeon RX 480 AOTS benchmark controversy…

 

Odd Radeon RX 480 AOTS Benchmark Results

In the final minutes of his presentation, Raja Koduri showed how two Radeon RX 480 graphics cards can beat the new NVIDIA GeForce GTX 1080 [Amazon] graphics card for far less money. Based on the sub-$500 price, the results are based on the 8 GB variant of the Radeon RX 480. Check it out in this short video clip :

What’s really odd is that Raja Koduri explicitly pointed out that the two Radeon RX 480 graphics cards were only utilised to 51% to slightly best the GeForce GTX 1080. Those are frankly, really odd numbers.

If it’s true that the two Radeon RX 480 graphics cards have a 49% headroom after beating the GeForce GTX 1080, then a single Radeon RX 480 would actually beat the GeForce GTX 1080 [Amazon]!

In fact, if we extrapolate the results so that both cards have the same 98.7% GPU utilisation, the Radeon RX 480 would deliver a frame rate of 60.5 fps. That would make the Radeon RX 480 about 3% faster than the GeForce GTX 1080 [Amazon], which we know is just not possible…

 

How The AOTS Benchmark Was Conducted

Robert Hallock, Head of Global Technical Marketing, AMD, helped to clarify the controversial Radeon RX 480 AOTS results. First, let’s start with the technical details of the benchmark setup :

Testbed System Specifications

CPU : Intel Core i7-5930K
Motherboard : ASRock X99M Killer
RAM : 32 GB DDR4-2400
Operating System : Microsoft Windows 10 (64-bit)

AMD GPU Configuration : 2 x Radeon RX 480 @ PCIe 3.0 x 16 for each GPU
AMD Driver : 16.30-160525n-230356E

NVIDIA GPU Configuration : GeForce GTX 1080 Founders Edition
NVIDIA Driver : 368.19

Ashes of the Singularity Version v1.12.19928
Ashes of the Singularity Game Settings : Crazy Settings | 1080P | 8x MSAA | VSYNC OFF

Robert then revealed the actual AOTS benchmark results and his take on them, which we will post verbatim :

Benchmark results:

2 x Radeon RX 480 – 62.5 fps | Single Batch GPU Util: 51% | Med Batch GPU Util: 71.9 | Heavy Batch GPU Util: 92.3%

GTX 1080 – 58.7 fps | Single Batch GPU Util: 98.7%| Med Batch GPU Util: 97.9% | Heavy Batch GPU Util: 98.7%

The elephant in the room:

Ashes uses procedural generation based on a randomized seed at launch. The benchmark does look slightly different every time it is run. But that, many have noted, does not fully explain the quality difference people noticed.

At present the GTX 1080 is incorrectly executing the terrain shaders responsible for populating the environment with the appropriate amount of snow. The GTX 1080 is doing less work to render AOTS than it otherwise would if the shader were being run properly. Snow is somewhat flat and boring in color compared to shiny rocks, which gives the illusion that less is being rendered, but this is an incorrect interpretation of how the terrain shaders are functioning in this title.

The content being rendered by the RX 480 — the one with greater snow coverage in the side-by-side (the left in these images) — is the correct execution of the terrain shaders.

So, even with fudgy image quality on the GTX 1080 that could improve their performance a few percent, dual RX 480 still came out ahead.

As a parting note, I will mention we ran this test 10x prior to going on-stage to confirm the performance delta was accurate. Moving up to 1440p at the same settings maintains the same performance delta within +/-1%.

* Credit for this find goes to Acid 에이스. Thanks, mate!

Next Page > The Odd Results Explained, Summary Of Key Points, What Does This Mean?

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participate in the Tech ARP Forums, or even donate to our fund. Any help you can render is greatly appreciated!

The Odd Radeon RX 480 AOTS Results Explained…

Robert Hallock was then asked about the odd 51% GPU utilisation result. He clarified that it was based on the single-batch (normal batch) data set, which was CPU-limited.

The single-batch or normal batch data set is about 60 seconds of scenes with lower (typically under 10,000) draw call counts. The medium batch data set consists of another 60 seconds of scenes with typical draw call counts between 10,000 and 20,000. The final heavy batch data set tests scenes with draw call counts in excess of 20,000.

To clarify this, the scaling from 1->2 GPUs in the dual RX 480 test we assembled is 1.83x. The OP was looking only at the lowest draw call rates when asking about the 51%. The single batch GPU utilization is 51% (CPU-bound), medium is 71.9% utilization (less CPU-bound) and heavy batch utilization is 92.3% (not CPU-bound). All together for the entire test, there is 1.83X the performance of a single GPU in what users saw on YouTube. The mGPU subsystem of AOTS is very robust.

 

Let Us Summarise The Points…

Based on his explanation, we can derive the following conclusions :

  • AMD claims that the GeForce GTX 1080 is performing a few percent better than it really should because it was improperly rendering the terrain in AOTS.
  • Two Radeon RX 480 graphics cards will deliver 83% better performance than a single Radeon RX 480 graphics card.
  • The benchmark scores are the average for the entire AOTS benchmark, which consists of the Normal Batch, Medium Batch and Heavy Batch data sets.
  • The 51% GPU utilisation was cherry-picked from the Normal Batch / Single Batch results.
  • The frame rate for Normal Batch / Single Batch data set was not revealed, but it would have been much, much lower than the 62.5 fps average frame rate.
  • The average GPU utilisation for the two AMD Radeon RX 480 cards was 71.7%, while the average GPU utilisation for the NVIDIA GeForce GTX 1080 [Amazon] was 98.4%.

In short, the controversy was created by AMD cherry-picking the GPU utilisation of the most CPU-limited portion of the AOTS benchmark, while comparing the average frame rate achieved in the entire AOTS benchmark.

[adrotate banner=”5″]

 

What Does This Mean?

Despite what Raja Koduri said in the video clip, there isn’t “much, much more headroom” for developers to squeeze out of the Radeon RX 480. That 49% headroom was only seen because the test was CPU-limited.

The two Radeon RX 480 cards were literally “chilling” 49% of the time, because they were waiting for the CPU to finish processing the AI and physics, before rendering a new frame. In fact, they most likely delivered a much lower frame rate as a result of being CPU-limited.

AMD inaccurately compared the average frame rate of the entire Ashes of the Singularity benchmark with the GPU utilisation of the Single Batch / Normal Batch portion of the benchmark (which is CPU-limited). Here is our rough correction of their comparison, based on the average GPU utilisation for the entire benchmark :

The adjusted results are still impressive, because it shows that there is still some headroom. Just not as much as was presented. We have no idea why they needed to cherry-pick that result, since the Radeon RX 480 will still impress gamers from the performance/cost point of view.

Based on our extrapolation of their comments, we can guesstimate that a single AMD Radeon RX 480 graphics card will deliver an average frame rate of 34 fps in the Ashes of the Singularity benchmark. That means the NVIDIA GeForce GTX 1080 Founders Edition [Amazon] is about 72.6% faster than the Radeon RX 480.

That sounds about right, and is hardly surprising since the GeForce GTX 1080 Founders Edition [Amazon] costs US$699 while the Radeon RX 480 (8 GB) will only cost US$229. There is nothing for AMD to be ashamed of achieving “only 58%” of the GeForce GTX 1080 Founders Edition’s performance. After all, it costs less than 1/3 of the price!

 

Support Tech ARP!

If you like our work, you can help support our work by visiting our sponsors, participate in the Tech ARP Forums, or even donate to our fund. Any help you can render is greatly appreciated!