Find out what AMD Smart Access Memory is all about, and how to turn it on for a FREE BOOST in performance!
Smart Access Memory : PCIe Resizable BAR for AMD!
Smart Access Memory is AMD’s marketing term for their implementation of the PCI Express Resizable BAR (Base Address Registers) capability.
What does that mean exactly?
CPUs are traditionally limited to a 256 MB I/O memory address region for the GPU frame buffer. This of it as an “data dump” for stuff like textures, shaders and geometry.
Since this “data dump” is limited to 256 MB, the CPU can only send texture, shader and geometry data as and when the GPU requires them.
This introduces some latency – delay from when the GPU requires the data, and the CPU send them.
Turning on Resizable BAR or Smart Access Memory greatly expands the size of that data dump, letting the CPU directly access the GPU’s entire frame buffer memory.
Instead of transferring data when requested by the GPU, the CPU processes and stores the data directly in the graphics memory.
Graphics assets can be transferred to graphics memory in full, instead of in pieces. In addition, multiple transfers can occur simultaneously, instead of being queued up.
While this AMD graphic above suggests that Smart Access Memory will widen the memory path (and thus memory bandwidth) between the CPU and GPU, that is not true.
Smart Access Memory / Resizable BAR will not increase memory bandwidth.
What it does is let the CPU directly access the entire GPU frame buffer memory, instead of using the usual 256 MB “dump”. That reduces latency because the graphics assets are now accessible by the GPU at all times.
AMD Smart Access Memory : Performance Gains
According to AMD, enabling Smart Access Memory will give you a small but free boost of 5% to 11% in gaming performance.
You can expect up to 16% better performance in some games, but no effect in certain games. But overall, you get a free boost in performance. There is simply no reason not to enable Smart Access Memory.
1080p Resolution (1920 x 1080)
1440p Resolution (2560 x 1440)
2160p Resolution (3840 x 2160)
AMD Smart Access Memory : Requirements
Since Smart Access Memory is just an AMD implementation of PCI Express Resizable BAR. Therefore, it can be be implemented for all PCI Express 3.0 and PCI Express 4.0 graphics cards and motherboards.
NVIDIA just announced that they will introduce Resizable BAR support at the end of February 2021!
Find out what Resizable BAR is all about, and why it matters!
Resizable BAR : What Is It?
Resizable BAR is an optional PCI Express feature, that can deliver a small but free boost in performance for the graphics card.
CPUs are traditionally limited to a 256 MB I/O memory address region for the GPU frame buffer. The CPU can only transfer data like textures, shaders and geometry to the GPU through that small 256 MB “window”.
Turning on Resizable BAR expands that small access window, letting the CPU directly access the GPU’s entire frame buffer memory.
Those graphics assets can thus be sent in full, instead of in pieces. In addition, multiple transfers can occur simultaneously, instead of being queued up.
NVIDIA To Introduce Resizable BAR In February 2021!
AMD was first out the door with Resizable BAR in November 2020, launching it as Smart Access Memory.
It gave their Radeon RX 6800 XT graphics card a free performance boost of up to 16% in some games, but no effect in other games.
On the other hand, AMD reused the chiplet design, with one or two CCDs (fabricated on 7 nm) paired with a 12 nm IOD (I/O Die).
Reads from CCD to IO are still 2X write, to conserve die area and transistor budget. And it uses the same IOD from Matisse (Zen 2).
AMD Zen 3 Transistor Count + Die Size
The new Zen 3 CCD has 4.15 billion transistors, with a die size of 80.7 mm². That’s up from the 3.8 billion transistors and a die size of 74 mm² for the Zen 2 CCD.
The Matisse-era IOD remains the same – 2.09 billion transistors, with a die size of 125 mm².
They will both be manufactured using the same 7 nm TSMC process for CCD, and 12 nm Global Foundries process for IOD.
7 nm TSMC
12 nm GoFlo
AMD Zen 3 Precision Boost : No Change
Precision Boost 2 in Zen 3 remains the same as that of Zen 2, just with higher frequencies to “play with”.
It is an opportunistic boost algorithm that drives the loaded cores to the highest possible frequency, until it hits any one of these limits :
VRM thermal limit
VRM current limit
maximum clock speed
Precision Boost 2 will dynamically analyse and boost or dither the core clock speeds every 1 ms using the Infinity Fabric command and control functions.
In this example of the new Ryzen 9 5900X with a base clock of 3.7 GHz and a boost clock of 4.8 GHz, Precision Boost 2 will typically result in frequencies of 4.4 GHz to 4.6 GHz, even with 24 threads running at the same time.
AMD Zen 3 Voltage Range : No Change
Despite the changes in the microarchitecture and SoC design, Zen 3 processors will continue to be engineered with the same voltage range as Zen- and Zen 2-based processors.
The typical voltage range will vary according to usage, but basically, Zen 3 processors, like the Ryzen 5000 series, will support 0.2 V to 1.5 V.
AMD Zen 3 Temperature Range : No Change
AMD Zen 3-based processors will also have the same temperature ranges as Zen- and Zen 2-based processors.
Note : The temperature range below assumes an enclosed chassis, and an air-conditioned room.