NVIDIA GeForce GTX 1660 Ti Tech Briefing + Q&A Session!

Spread the love

NVIDIA TU116, GDDR6, Power, Concurrent FP + Int Ops


The NVIDIA GeForce GTX 1660 Ti uses a brand new TU116 GPU that is built using the latest NVIDIA Turing microarchitecture.

However, it lacks the Tensor and RT cores that GeForce RTX graphics cards have. This is the reason why NVIDIA pegs It as a GTX, and not RTX.

The Official NVIDIA GeForce GTX 1660 Ti Tech Briefing

The NVIDIA TU116 GPU is not a cut-down version of the TU106, but a completely new design with the Tensor cores and RT cores removed. It retains the other advantages of the Turing architecture – Concurrent Floating Point + Integer Operations, a Unified Cache Architecture and Adaptive Shading.

The Official NVIDIA GeForce GTX 1660 Ti Tech Briefing

The TU116 basically bridges the old Pascal-based GeForce GTX graphics cards and the new Turing-based GeForce RTX graphics cards. Even though it lacks RT and Tensor cores, its Turing shaders gives it a big performance advantage.


GDDR6 Memory

The NVIDIA GeForce GTX 1660 Ti comes with 6 GB of the GDDR6 memory. Coupled with its 192-bit memory interface, it is able to deliver a peak memory bandwidth of 288.1 GB/s. This is 50% more than the memory bandwidth of the GeForce GTX 1060.


Power Usage

The NVIDIA TU116 has 6.6 billion transistors – 50% more than the GP106 GPU used in the GeForce GTX 1060. But thanks to its smaller 12 nm FinFet process technology, NVIDIA is able to keep its thermal output and power consumption low.

Like the GeForce GTX 1060, the NVIDIA GeForce GTX 1660 Ti has a 120 watt TDP, and requires one 8-pin PCIe power connector.


Concurrent Floating Point + Integer Operations

This is an important Turing shader technology, because games are increasingly mixing floating point operations with integer instructions. Take, for example, Shadow of the Tomb Raider.

For every 100 instructions, there are on average 62 floating point and 32 integer instructions. In earlier NVIDIA GPU architectures, the floating point datapath would idle while integer instructions are running.

The Official NVIDIA GeForce GTX 1660 Ti Tech Briefing

[adrotate group=”1″]

The NVIDIA TU116 features a second, parallel integer execution unit. This allows it to execute integer instructions in parallel with floating point calculations. With an ideal 50-50 mix, the GeForce GTX 1660 Ti has more than twice the processing power of the GeForce GTX 1060

The Official NVIDIA GeForce GTX 1660 Ti Tech Briefing

Of course, in real life, the actual performance boost is lower. According to NVIDIA, this feature alone gives the GeForce GTX 1660 Ti a 50% performance advantage over the GeForce GTX 1060.

The Official NVIDIA GeForce GTX 1660 Ti Tech Briefing

Next Page > Unified Cache Architecture + Adaptive Shading


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!

About The Author

Dr. Adrian Wong has been writing about tech and science since 1997, even publishing a book with Prentice Hall called Breaking Through The BIOS Barrier (ISBN 978-0131455368) while in medical school. He continues to devote countless hours every day writing about tech, medicine and science, in his pursuit of facts in a post-truth world.

Leave a Reply