NVIDIA A100 PCIe vs NVIDIA A16 PCIe
Graphics Processor
| GPU model | GA100 | GA107 |
|---|---|---|
| Architecture | Ampere | Ampere |
| Foundry | TSMC | Samsung |
| Manufacturing process | 7 nm | 8 nm |
| Number of transistors | 54,200 million | unknown |
| Die size | 826 mm² | unknown |
Graphics Card
| Release date | Jun 22nd, 2020 | Apr 12th, 2021 |
|---|---|---|
| Generation | Tesla | Tesla |
| Production | Active | Active |
| Interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
| gpu.details.availability | — | 2021 |
Clocks
| GPU clock | 1410 MHz | — |
|---|---|---|
| Memory clock | 1215 MHz 2.4 Gbps effective | 1812 MHz 14.5 Gbps effective |
| Base clock | 765 MHz | 885 MHz |
| Boost clock | 1410 MHz | 1695 MHz |
Memory Configuration
| Memory size | 40 GB | 16 GB |
|---|---|---|
| Memory type | HBM2E | GDDR6 |
| Memory bus width | 5120 bit | 128 bit |
| Bandwidth | 1,555 GB/s | 231.9 GB/s |
Render Configuration
| Shading units | 6912 | 2560 |
|---|---|---|
| TMUs | 432 | 80 |
| ROPs | 160 | 48 |
| SM count | 108 | 20 |
| Tensor cores | 432 | 80 |
| Cache L1 | 192 KB (per SM) | 128 KB (per SM) |
| Cache L2 | 40 MB | 2 MB |
| RT cores | — | 20 |
Performance
| Pixel rate | 225.6 GPixel/s | 81.36 GPixel/s |
|---|---|---|
| Texture rate | 609.1 GTexel/s | 135.6 GTexel/s |
| FP16 (half) performance | 77.97 TFLOPS (4:1) | 8.678 TFLOPS (1:1) |
| FP32 (float) performance | 19.49 TFLOPS | 8.678 TFLOPS |
| FP64 (double) performance | 9.746 TFLOPS (1:2) | 271.2 GFLOPS (1:32) |
Dimensions & Outputs
| Slot width | Dual-slot | Dual-slot |
|---|---|---|
| Length | 267 mm 10.5 inches | 267 mm 10.5 inches |
| TDP | 250 W | 250 W |
| Suggested PSU | 600 W | 600 W |
| Outputs | No outputs | No outputs |
| Power connectors | 8-pin EPS | 8-pin EPS |
| Board number | P1001 SKU 200 | — |
| Width | — | 112 mm 4.4 inches |
API Support & Features
| DirectX | — | 12 Ultimate (12_2) |
|---|---|---|
| OpenGL | — | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | — | 1.2 |
| CUDA | 8.0 | 8.6 |
| Shader model | — | 6.6 |
