GPU model | Polaris 22 | TU117 |
---|---|---|
GPU variant | VegaM XL (D715A714) | N18S-G5 |
Architecture | GCN 4.0 | Turing |
Foundry | GlobalFoundries | TSMC |
Manufacturing process | 14 nm | 12 nm |
Number of transistors | 5,000 million | 4,700 million |
Die size | 208 mm² | 200 mm² |
Release date | Feb 1st, 2018 | Aug 25th, 2020 |
---|---|---|
Generation | Vega (Vega M) | GeForce MX (4xx) |
Production | Active | Active |
Interface | IGP | PCIe 4.0 x4 |
Base clock | 931 MHz | 1395 MHz |
---|---|---|
Boost clock | 1011 MHz | 1575 MHz |
Memory clock | 700 MHz 1400 Mbps effective | 1250 MHz 10 Gbps effective |
Memory size | 4 GB | 2 GB |
---|---|---|
Memory type | HBM2 | GDDR6 |
Memory bus width | 1024 bit | 64 bit |
Bandwidth | 179.2 GB/s | 80.00 GB/s |
Shading units | 1280 | 896 |
---|---|---|
TMUs | 80 | 56 |
ROPs | 32 | 32 |
Compute units | 20 | — |
Cache L1 | 16 KB (per CU) | 64 KB (per SM) |
Cache L2 | 1024 KB | 512 KB |
SM count | — | 14 |
Pixel rate | 32.35 GPixel/s | 50.40 GPixel/s |
---|---|---|
Texture rate | 80.88 GTexel/s | 88.20 GTexel/s |
FP16 (half) performance | 2.588 TFLOPS (1:1) | 5.645 TFLOPS (2:1) |
FP32 (float) performance | 2.588 TFLOPS | 2.822 TFLOPS |
FP64 (double) performance | 161.8 GFLOPS (1:16) | 88.20 GFLOPS (1:32) |
Slot width | IGP | — |
---|---|---|
TDP | 65 W | 31 W |
Outputs | No outputs | No outputs |
Board number | D136 | — |
Power connectors | — | None |
DirectX | 12 (12_0) | 12 (12_1) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 2.1 | 3.0 |
Vulkan | 1.2 | 1.2 |
Shader model | 6.4 | 6.6 |
CUDA | — | 7.5 |