GPU model | Polaris 22 | TU117B |
---|---|---|
GPU variant | VegaM XT (C7398773) | N19P-Q1-A1 |
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 | Jun 8th, 2020 |
---|---|---|
Generation | Vega (Vega M) | Quadro Mobile (Tx000) |
Production | Active | Active |
Interface | IGP | PCIe 3.0 x16 |
Base clock | 1063 MHz | 1395 MHz |
---|---|---|
Boost clock | 1190 MHz | 1650 MHz |
Memory clock | 800 MHz 1600 Mbps effective | 1500 MHz 12 Gbps effective |
Memory size | 4 GB | 4 GB |
---|---|---|
Memory type | HBM2 | GDDR6 |
Memory bus width | 1024 bit | 128 bit |
Bandwidth | 204.8 GB/s | 192.0 GB/s |
Shading units | 1536 | 896 |
---|---|---|
TMUs | 96 | 56 |
ROPs | 64 | 32 |
Compute units | 24 | — |
Cache L1 | 16 KB (per CU) | 64 KB (per SM) |
Cache L2 | 1024 KB | 1024 KB |
SM count | — | 14 |
Pixel rate | 76.16 GPixel/s | 52.80 GPixel/s |
---|---|---|
Texture rate | 114.2 GTexel/s | 92.40 GTexel/s |
FP16 (half) performance | 3.656 TFLOPS (1:1) | 5.914 TFLOPS (2:1) |
FP32 (float) performance | 3.656 TFLOPS | 2.957 TFLOPS |
FP64 (double) performance | 228.5 GFLOPS (1:16) | 92.40 GFLOPS (1:32) |
Slot width | IGP | IGP |
---|---|---|
TDP | 100 W | 50 W |
Outputs | No outputs | No outputs |
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 |