ATI Radeon HD 2900 XT vs NVIDIA GRID A100B
Graphics Processor
| GPU model | R600 | GA100 |
|---|---|---|
| GPU variant | R600 XT (215RGMDKA13FG) | — |
| Architecture | TeraScale | Ampere |
| Foundry | TSMC | TSMC |
| Manufacturing process | 80 nm | 7 nm |
| Number of transistors | 720 million | 54,200 million |
| Die size | 420 mm² | 826 mm² |
Graphics Card
| Release date | May 14th, 2007 | May 14th, 2020 |
|---|---|---|
| Generation | Radeon R600 | GRID |
| Production | End-of-life | Active |
| Launch price | 399 USD | — |
| Interface | PCIe 1.0 x16 | PCIe 4.0 x16 |
| Reviews | 63 in our database | — |
| Predecessor | Radeon R500 PCIe | — |
| Successor | Radeon R700 | — |
Clocks
| GPU clock | 743 MHz | — |
|---|---|---|
| Memory clock | 828 MHz 1656 Mbps effective | 1215 MHz 2.4 Gbps effective |
| Base clock | — | 900 MHz |
| Boost clock | — | 1005 MHz |
Memory Configuration
| Memory size | 512 MB | 48 GB |
|---|---|---|
| Memory type | GDDR3 | HBM2E |
| Memory bus width | 512 bit | 6144 bit |
| Bandwidth | 106.0 GB/s | 1,866 GB/s |
Render Configuration
| Shading units | 320 | 6912 |
|---|---|---|
| TMUs | 16 | 432 |
| ROPs | 16 | 192 |
| Compute units | 4 | — |
| Cache L2 | 256 KB | 48 MB |
| SM count | — | 108 |
| Tensor cores | — | 432 |
| Cache L1 | — | 192 KB (per SM) |
Performance
| Pixel rate | 11.89 GPixel/s | 193.0 GPixel/s |
|---|---|---|
| Texture rate | 11.89 GTexel/s | 434.2 GTexel/s |
| FP32 (float) performance | 475.5 GFLOPS | 13.89 TFLOPS |
| FP16 (half) performance | — | 55.57 TFLOPS (4:1) |
| FP64 (double) performance | — | 6.947 TFLOPS (1:2) |
Dimensions & Outputs
| Slot width | Dual-slot | IGP |
|---|---|---|
| Length | 241 mm 9.5 inches | — |
| TDP | 215 W | 400 W |
| Suggested PSU | 550 W | 800 W |
| Outputs | 2x DVI1x S-Video | No outputs |
| Power connectors | 1x 6-pin + 1x 8-pin | None |
| Board number | B006-01 | — |
API Support & Features
| DirectX | 10.0 (10_0) | — |
|---|---|---|
| OpenGL | 3.3 (full) 4.0 (partial) | — |
| OpenCL | — | 3.0 |
| Vulkan | — | — |
| Shader model | 4.0 | — |
| CUDA | — | 8.0 |
