GPU Name | Navi 23 | TU102 |
---|---|---|
Architektur | RDNA 2.0 | Turing |
Hersteller | TSMC | TSMC |
Prozessgröße | 7 nm | 12 nm |
Transistoren | 11,060 million | 18,600 million |
Chipgröße | 237 mm² | 754 mm² |
GPU Variante | — | TU102-875-A1 |
Veröffentlichungsdatum | Jun 8th, 2021 | Aug 13th, 2018 |
---|---|---|
gpu.details.availability | Jul, 2021 | — |
Generation | Radeon Pro | Quadro |
Produktion | Active | Active |
Originalpreis | 649 USD | 6,299 USD |
Businterface | PCIe 4.0 x16 | PCIe 3.0 x16 |
Basis-Taktfrequenz | 2331 MHz | 1305 MHz |
---|---|---|
Boost-Takt | 2903 MHz | 1560 MHz |
Speicher-Taktung | 1750 MHz 14 Gbps effective | 1750 MHz 14 Gbps effective |
Speichergröße | 8 GB | 24 GB |
---|---|---|
Speichertyp | GDDR6 | GDDR6 |
Speicherbus | 128 bit | 384 bit |
Bandbreite | 224.0 GB/s | 672.0 GB/s |
Shading-Einheiten | 1792 | 4608 |
---|---|---|
TMUs | 112 | 288 |
ROPs | 64 | 96 |
Compute Units | 28 | — |
RT-Kerne | 28 | 72 |
gpu.details.l0-cache | 32 KB per WGP | — |
Cache L1 | 128 KB per Array | 64 KB (per SM) |
Cache L2 | 2 MB | 6 MB |
gpu.details.l3-cache | 32 MB | — |
Anzahl SM | — | 72 |
Tensor Cores | — | 576 |
Pixelrate | 185.8 GPixel/s | 149.8 GPixel/s |
---|---|---|
Texturrate | 325.1 GTexel/s | 449.3 GTexel/s |
FP16 (half) Leistung | 20.81 TFLOPS (2:1) | 28.75 TFLOPS (2:1) |
FP32 (float) Leistung | 10.40 TFLOPS | 14.38 TFLOPS |
FP64 (double) Leistung | 650.3 GFLOPS (1:16) | 449.3 GFLOPS (1:32) |
Steckplatzbreite | Single-slot | Dual-slot |
---|---|---|
Länge | 241 mm 9.5 inches | 267 mm 10.5 inches |
TDP | 100 W | 260 W |
Vorgeschlagenes Netzteil | 300 W | 600 W |
Ausgabe | 4x DisplayPort | No outputs |
Stromverbindung | 1x 6-pin | 1x 6-pin + 1x 8-pin |
Breite | — | 111 mm 4.4 inches |
Board-Nummer | — | PG150 SKU 510 |
DirectX | 12.0 Ultimate (12_2) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 2.1 | 3.0 |
Vulkan | 1.2 | 1.2 |
Shader-Modell | 6.5 | 6.6 |
CUDA | — | 7.5 |