GPU model | Baffin | GA100 |
---|---|---|
GPU variant | Baffin LE | — |
Architecture | GCN 4.0 | Ampere |
Foundry | GlobalFoundries | TSMC |
Manufacturing process | 14 nm | 7 nm |
Number of transistors | 3,000 million | 54,200 million |
Die size | 123 mm² | 826 mm² |
Release date | Oct 13th, 2017 | May 14th, 2020 |
---|---|---|
Generation | Polaris | GRID |
Production | Active | Active |
Interface | PCIe 3.0 x8 | PCIe 4.0 x16 |
Predecessor | Arctic Islands | — |
Successor | Vega | — |
Base clock | 1019 MHz | 900 MHz |
---|---|---|
Boost clock | 1071 MHz | 1005 MHz |
Memory clock | 1500 MHz 6 Gbps effective | 1215 MHz 2.4 Gbps effective |
Memory size | 2 GB | 48 GB |
---|---|---|
Memory type | GDDR5 | HBM2E |
Memory bus width | 128 bit | 6144 bit |
Bandwidth | 96.00 GB/s | 1,866 GB/s |
Shading units | 640 | 6912 |
---|---|---|
TMUs | 40 | 432 |
ROPs | 16 | 192 |
Compute units | 10 | — |
Cache L1 | 16 KB (per CU) | 192 KB (per SM) |
Cache L2 | 1024 KB | 48 MB |
SM count | — | 108 |
Tensor cores | — | 432 |
Pixel rate | 17.14 GPixel/s | 193.0 GPixel/s |
---|---|---|
Texture rate | 42.84 GTexel/s | 434.2 GTexel/s |
FP16 (half) performance | 1,371 GFLOPS (1:1) | 55.57 TFLOPS (4:1) |
FP32 (float) performance | 1,371 GFLOPS | 13.89 TFLOPS |
FP64 (double) performance | 85.68 GFLOPS (1:16) | 6.947 TFLOPS (1:2) |
Slot width | Dual-slot | IGP |
---|---|---|
Length | 145 mm 5.7 inches | — |
TDP | 60 W | 400 W |
Suggested PSU | 250 W | 800 W |
Outputs | 1x DVI1x HDMI1x DisplayPort | No outputs |
Power connectors | None | None |
Board number | D090-21 | — |
DirectX | 12 (12_0) | — |
---|---|---|
OpenGL | 4.6 | — |
OpenCL | 2.1 | 3.0 |
Vulkan | 1.2 | — |
Shader model | 6.4 | — |
CUDA | — | 8.0 |