AMD Radeon Pro Vega II vs NVIDIA Quadro RTX 4000
Grafikprozessor
| GPU Name | Vega 20 | TU104 |
|---|---|---|
| GPU Variante | Vega 20 XT | TU104-850-A1 |
| Architektur | GCN 5.1 | Turing |
| Hersteller | TSMC | TSMC |
| Prozessgröße | 7 nm | 12 nm |
| Transistoren | 13,230 million | 13,600 million |
| Chipgröße | 331 mm² | 545 mm² |
Grafikkarte
| Veröffentlichungsdatum | Jun 3rd, 2019 | Nov 13th, 2018 |
|---|---|---|
| Generation | Radeon Pro Mac | Quadro |
| Produktion | Active | Active |
| Businterface | PCIe 3.0 x16 | PCIe 3.0 x16 |
| Originalpreis | — | 899 USD |
Takt-Geschwindigkeiten
| Basis-Taktfrequenz | 1574 MHz | 1005 MHz |
|---|---|---|
| Boost-Takt | 1720 MHz | 1545 MHz |
| Speicher-Taktung | 806 MHz 1612 Mbps effective | 1625 MHz 13 Gbps effective |
Speicher
| Speichergröße | 32 GB | 8 GB |
|---|---|---|
| Speichertyp | HBM2 | GDDR6 |
| Speicherbus | 4096 bit | 256 bit |
| Bandbreite | 825.3 GB/s | 416.0 GB/s |
Render-Konfiguration
| Shading-Einheiten | 4096 | 2304 |
|---|---|---|
| TMUs | 256 | 144 |
| ROPs | 64 | 64 |
| Compute Units | 64 | — |
| Cache L1 | 16 KB (per CU) | 64 KB (per SM) |
| Cache L2 | 4 MB | 4 MB |
| Anzahl SM | — | 36 |
| Tensor Cores | — | 288 |
| RT-Kerne | — | 36 |
Theoretische Leistung
| Pixelrate | 110.1 GPixel/s | 98.88 GPixel/s |
|---|---|---|
| Texturrate | 440.3 GTexel/s | 222.5 GTexel/s |
| FP16 (half) Leistung | 28.18 TFLOPS (2:1) | 14.24 TFLOPS (2:1) |
| FP32 (float) Leistung | 14.09 TFLOPS | 7.119 TFLOPS |
| FP64 (double) Leistung | 880.6 GFLOPS (1:16) | 222.5 GFLOPS (1:32) |
Boarddesign
| Steckplatzbreite | Quad-slot | Single-slot |
|---|---|---|
| TDP | 475 W | 160 W |
| Vorgeschlagenes Netzteil | 850 W | 450 W |
| Ausgabe | 1x HDMI4x Thunderbolt | 3x DisplayPort1x USB Type-C |
| Stromverbindung | None | 1x 8-pin |
| Board-Nummer | MW672ZM/A | — |
| Länge | — | 241 mm 9.5 inches |
| Breite | — | 111 mm 4.4 inches |
Grafik-Features
| DirectX | 12 (12_1) | 12 Ultimate (12_2) |
|---|---|---|
| OpenGL | 4.6 | 4.6 |
| OpenCL | 2.1 | 3.0 |
| Vulkan | 1.2 | 1.2 |
| Shader-Modell | 6.4 | 6.6 |
| CUDA | — | 7.5 |
