NVIDIA CMP 40HX vs NVIDIA CMP 70HX
Graphics Processor
| GPU model | TU106 | GA104 |
|---|---|---|
| GPU variant | TU106-100-A1 | GA104-100-A1 |
| Architecture | Turing | Ampere |
| Foundry | TSMC | Samsung |
| Manufacturing process | 12 nm | 8 nm |
| Number of transistors | 10,800 million | 17,400 million |
| Die size | 445 mm² | 392 mm² |
Graphics Card
| Release date | Feb 25th, 2021 | Unknown |
|---|---|---|
| Generation | Mining GPUs | Mining GPUs |
| Production | Active | Active |
| Launch price | 699 USD | — |
| Interface | PCIe 3.0 x16 | PCIe 4.0 x16 |
Clocks
| Base clock | 1470 MHz | 1365 MHz |
|---|---|---|
| Boost clock | 1650 MHz | 1395 MHz |
| Memory clock | 1750 MHz 14 Gbps effective | 1188 MHz 19 Gbps effective |
Memory Configuration
| Memory size | 8 GB | 8 GB |
|---|---|---|
| Memory type | GDDR6 | GDDR6X |
| Memory bus width | 256 bit | 256 bit |
| Bandwidth | 448.0 GB/s | 608.3 GB/s |
Render Configuration
| Shading units | 2304 | 6144 |
|---|---|---|
| TMUs | 144 | 192 |
| ROPs | 64 | 96 |
| SM count | 36 | 48 |
| Tensor cores | 288 | 192 |
| RT cores | 36 | 48 |
| Cache L1 | 64 KB (per SM) | 128 KB (per SM) |
| Cache L2 | 4 MB | 4 MB |
Performance
| Pixel rate | 105.6 GPixel/s | 133.9 GPixel/s |
|---|---|---|
| Texture rate | 237.6 GTexel/s | 267.8 GTexel/s |
| FP16 (half) performance | 15.21 TFLOPS (2:1) | 17.14 TFLOPS (1:1) |
| FP32 (float) performance | 7.603 TFLOPS | 17.14 TFLOPS |
| FP64 (double) performance | 237.6 GFLOPS (1:32) | 267.8 GFLOPS (1:64) |
Dimensions & Outputs
| Slot width | Dual-slot | Dual-slot |
|---|---|---|
| Length | 229 mm 9 inches | 267 mm 10.5 inches |
| Width | 111 mm 4.4 inches | 112 mm 4.4 inches |
| Height | 35 mm 1.4 inches | — |
| TDP | 185 W | unknown |
| Suggested PSU | 450 W | 200 W |
| Outputs | No outputs | No outputs |
| Power connectors | 1x 8-pin | 1x 12-pin |
| Board number | PG161 SKU 100 | — |
API Support & Features
| DirectX | 12 Ultimate (12_2) | 12 Ultimate (12_2) |
|---|---|---|
| OpenGL | 4.6 | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.2 | 1.2 |
| CUDA | 7.5 | 8.6 |
| Shader model | 6.6 | 6.6 |
