ATI Radeon HD 4810 vs NVIDIA GRID RTX T10-4
Graphics Processor
| GPU model | RV770 | TU102 |
|---|---|---|
| GPU variant | RV770 CE (215-0669093) | TU102-875-A1 |
| Architecture | TeraScale | Turing |
| Foundry | TSMC | TSMC |
| Manufacturing process | 55 nm | 12 nm |
| Number of transistors | 956 million | 18,600 million |
| Die size | 256 mm² | 754 mm² |
Graphics Card
| Release date | May 28th, 2009 | Unknown |
|---|---|---|
| Generation | Radeon R700 | GRID |
| Production | End-of-life | Active |
| Interface | PCIe 2.0 x16 | PCIe 3.0 x16 |
| Predecessor | Radeon R600 | — |
| Successor | Evergreen | — |
Clocks
| GPU clock | 625 MHz | — |
|---|---|---|
| Memory clock | 900 MHz 3.6 Gbps effective | 1750 MHz 14 Gbps effective |
| Base clock | — | 1065 MHz |
| Boost clock | — | 1395 MHz |
Memory Configuration
| Memory size | 512 MB | 4 GB |
|---|---|---|
| Memory type | GDDR5 | GDDR6 |
| Memory bus width | 128 bit | 384 bit |
| Bandwidth | 57.60 GB/s | 672.0 GB/s |
Render Configuration
| Shading units | 640 | 4608 |
|---|---|---|
| TMUs | 32 | 288 |
| ROPs | 8 | 96 |
| Compute units | 8 | — |
| Cache L1 | 16 KB (per CU) | 64 KB (per SM) |
| Cache L2 | 128 KB | 6 MB |
| SM count | — | 72 |
| Tensor cores | — | 576 |
| RT cores | — | 72 |
Performance
| Pixel rate | 5.000 GPixel/s | 133.9 GPixel/s |
|---|---|---|
| Texture rate | 20.00 GTexel/s | 401.8 GTexel/s |
| FP32 (float) performance | 800.0 GFLOPS | 12.86 TFLOPS |
| FP64 (double) performance | 160.0 GFLOPS (1:5) | 401.8 GFLOPS (1:32) |
| FP16 (half) performance | — | 25.71 TFLOPS (2:1) |
Dimensions & Outputs
| Slot width | Single-slot | Dual-slot |
|---|---|---|
| Length | 246 mm 9.7 inches | 267 mm 10.5 inches |
| TDP | 95 W | 260 W |
| Suggested PSU | 250 W | 600 W |
| Outputs | 2x DVI1x S-Video | No outputs |
| Power connectors | 1x 6-pin | 1x 6-pin + 1x 8-pin |
| Board number | B507 | PG150 SKU 215 |
API Support & Features
| DirectX | 10.1 (10_1) | 12 Ultimate (12_2) |
|---|---|---|
| OpenGL | 3.3 | 4.6 |
| OpenCL | 1.1 | 3.0 |
| Vulkan | — | 1.2 |
| Shader model | 4.1 | 6.6 |
| CUDA | — | 7.5 |
