NVIDIA CMP 40HX vs NVIDIA P104-100
Grafikprozessor
| GPU Name | TU106 | GP104 |
|---|---|---|
| GPU Variante | TU106-100-A1 | GP104-100-A1 |
| Architektur | Turing | Pascal |
| Hersteller | TSMC | TSMC |
| Prozessgröße | 12 nm | 16 nm |
| Transistoren | 10,800 million | 7,200 million |
| Chipgröße | 445 mm² | 314 mm² |
Grafikkarte
| Veröffentlichungsdatum | Feb 25th, 2021 | Dec 12th, 2017 |
|---|---|---|
| Generation | Mining GPUs | Mining GPUs |
| Produktion | Active | End-of-life |
| Originalpreis | 699 USD | — |
| Businterface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Takt-Geschwindigkeiten
| Basis-Taktfrequenz | 1470 MHz | 1607 MHz |
|---|---|---|
| Boost-Takt | 1650 MHz | 1733 MHz |
| Speicher-Taktung | 1750 MHz 14 Gbps effective | 1251 MHz 10 Gbps effective |
Speicher
| Speichergröße | 8 GB | 4 GB |
|---|---|---|
| Speichertyp | GDDR6 | GDDR5X |
| Speicherbus | 256 bit | 256 bit |
| Bandbreite | 448.0 GB/s | 320.3 GB/s |
Render-Konfiguration
| Shading-Einheiten | 2304 | 1920 |
|---|---|---|
| TMUs | 144 | 120 |
| ROPs | 64 | 64 |
| Anzahl SM | 36 | 15 |
| Tensor Cores | 288 | — |
| RT-Kerne | 36 | — |
| Cache L1 | 64 KB (per SM) | 48 KB (per SM) |
| Cache L2 | 4 MB | 2 MB |
Theoretische Leistung
| Pixelrate | 105.6 GPixel/s | 110.9 GPixel/s |
|---|---|---|
| Texturrate | 237.6 GTexel/s | 208.0 GTexel/s |
| FP16 (half) Leistung | 15.21 TFLOPS (2:1) | 104.0 GFLOPS (1:64) |
| FP32 (float) Leistung | 7.603 TFLOPS | 6.655 TFLOPS |
| FP64 (double) Leistung | 237.6 GFLOPS (1:32) | 208.0 GFLOPS (1:32) |
Boarddesign
| Steckplatzbreite | Dual-slot | Dual-slot |
|---|---|---|
| Länge | 229 mm 9 inches | 267 mm 10.5 inches |
| Breite | 111 mm 4.4 inches | — |
| Höhe | 35 mm 1.4 inches | — |
| TDP | 185 W | unknown |
| Vorgeschlagenes Netzteil | 450 W | 200 W |
| Ausgabe | No outputs | No outputs |
| Stromverbindung | 1x 8-pin | 1x 8-pin |
| Board-Nummer | PG161 SKU 100 | PG413 SKU 100 |
Grafik-Features
| DirectX | 12 Ultimate (12_2) | 12 (12_1) |
|---|---|---|
| OpenGL | 4.6 | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.2 | 1.2 |
| CUDA | 7.5 | 6.1 |
| Shader-Modell | 6.6 | 6.4 |
