AMD Radeon Instinct MI100 vs NVIDIA Quadro P1000
Grafikprozessor
| GPU Name | Arcturus | GP107 |
|---|---|---|
| GPU Variante | Arcturus XL | GP107-860-A1 |
| Architektur | CDNA 1.0 | Pascal |
| Hersteller | TSMC | Samsung |
| Prozessgröße | 7 nm | 14 nm |
| Transistoren | 25,600 million | 3,300 million |
| Chipgröße | 750 mm² | 132 mm² |
Grafikkarte
| Veröffentlichungsdatum | Nov 16th, 2020 | Feb 7th, 2017 |
|---|---|---|
| Generation | Radeon Instinct | Quadro |
| Produktion | Active | End-of-life |
| Businterface | PCIe 4.0 x16 | PCIe 3.0 x16 |
Takt-Geschwindigkeiten
| Basis-Taktfrequenz | 1000 MHz | 1354 MHz |
|---|---|---|
| Boost-Takt | 1502 MHz | 1392 MHz |
| SOC-Taktung | 1091 MHz | — |
| UVD-Taktung | 1403 MHz | — |
| Speicher-Taktung | 1200 MHz 2.4 Gbps effective | 1502 MHz 6 Gbps effective |
Speicher
| Speichergröße | 32 GB | 4 GB |
|---|---|---|
| Speichertyp | HBM2 | GDDR5 |
| Speicherbus | 4096 bit | 128 bit |
| Bandbreite | 1,229 GB/s | 96.13 GB/s |
Render-Konfiguration
| Shading-Einheiten | 7680 | 512 |
|---|---|---|
| TMUs | 480 | 32 |
| ROPs | 64 | 16 |
| Compute Units | 120 | — |
| Cache L1 | 16 KB (per CU) | 48 KB (per SM) |
| Cache L2 | 8 MB | 1024 KB |
| Anzahl SM | — | 4 |
Theoretische Leistung
| Pixelrate | 96.13 GPixel/s | 22.27 GPixel/s |
|---|---|---|
| Texturrate | 721.0 GTexel/s | 44.54 GTexel/s |
| FP16 (half) Leistung | 184.6 TFLOPS (8:1) | 22.27 GFLOPS (1:64) |
| FP32 (float) Leistung | 23.07 TFLOPS | 1,425 GFLOPS |
| FP64 (double) Leistung | 11.54 TFLOPS (1:2) | 44.54 GFLOPS (1:32) |
Boarddesign
| Steckplatzbreite | Dual-slot | Single-slot |
|---|---|---|
| Länge | 267 mm 10.5 inches | 150 mm 5.9 inches |
| Breite | 111 mm 4.4 inches | 69 mm 2.7 inches |
| TDP | 300 W | 47 W |
| Vorgeschlagenes Netzteil | 700 W | 200 W |
| Ausgabe | No outputs | 4x mini-DisplayPort |
| Stromverbindung | 2x 8-pin | None |
| Board-Nummer | 109-D05187 | PG212 SKU 502 |
| gpu.details.bios-number | 113-D3430300-025 | — |
Grafik-Features
| DirectX | — | 12 (12_1) |
|---|---|---|
| OpenGL | — | 4.6 |
| OpenCL | 2.1 | 3.0 |
| Vulkan | — | 1.2 |
| Shader-Modell | — | 6.4 |
| CUDA | — | 6.1 |
