GPU model | Vega 20 | GP100 |
---|---|---|
GPU variant | Vega 20 XT | — |
Architecture | GCN 5.1 | Pascal |
Foundry | TSMC | TSMC |
Manufacturing process | 7 nm | 16 nm |
Number of transistors | 13,230 million | 15,300 million |
Die size | 331 mm² | 610 mm² |
Release date | Jun 3rd, 2019 | Oct 1st, 2016 |
---|---|---|
Generation | Radeon Pro Mac | Quadro |
Production | Active | End-of-life |
Interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Base clock | 1574 MHz | 1304 MHz |
---|---|---|
Boost clock | 1720 MHz | 1442 MHz |
Memory clock | 806 MHz 1612 Mbps effective | 715 MHz 1430 Mbps effective |
Memory size | 32 GB | 16 GB |
---|---|---|
Memory type | HBM2 | HBM2 |
Memory bus width | 4096 bit | 4096 bit |
Bandwidth | 825.3 GB/s | 732.2 GB/s |
Shading units | 4096 | 3584 |
---|---|---|
TMUs | 256 | 224 |
ROPs | 64 | 96 |
Compute units | 64 | — |
Cache L1 | 16 KB (per CU) | 24 KB (per SM) |
Cache L2 | 4 MB | 4 MB |
SM count | — | 56 |
Pixel rate | 110.1 GPixel/s | 138.4 GPixel/s |
---|---|---|
Texture rate | 440.3 GTexel/s | 323.0 GTexel/s |
FP16 (half) performance | 28.18 TFLOPS (2:1) | 20.67 TFLOPS (2:1) |
FP32 (float) performance | 14.09 TFLOPS | 10.34 TFLOPS |
FP64 (double) performance | 880.6 GFLOPS (1:16) | 5.168 TFLOPS (1:2) |
Slot width | Quad-slot | Dual-slot |
---|---|---|
TDP | 475 W | 235 W |
Suggested PSU | 850 W | 550 W |
Outputs | 1x HDMI4x Thunderbolt | 1x DVI4x DisplayPort |
Power connectors | None | 1x 8-pin |
Board number | MW672ZM/A | — |
Length | — | 267 mm 10.5 inches |
Width | — | 111 mm 4.4 inches |
DirectX | 12 (12_1) | 12 (12_1) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 2.1 | 3.0 |
Vulkan | 1.2 | 1.2 |
Shader model | 6.4 | 6.4 |
CUDA | — | 6.0 |