Nom du GPU | Navi 10 | GA104 |
---|---|---|
GPU variant | Navi 10 XLE (215-0917338) | GA104-200-A1 |
Architecture | RDNA 1.0 | Ampere |
Fonderie | TSMC | Samsung |
Taille du processus | 7 nm | 8 nm |
Transistors | 10,300 million | 17,400 million |
Taille des matrices | 251 mm² | 392 mm² |
Date de publication | Jan 21st, 2020 | Dec 1st, 2020 |
---|---|---|
Génération | Navi | GeForce 30 |
Prédécesseur | Vega | GeForce 20 |
Production | Active | Active |
Prix de lancement | 279 USD | 399 USD |
Bus interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
Critiques | 78 in our database | 64 in our database |
Successeur | Navi II | — |
gpu.details.availability | — | Dec 2nd, 2020 |
Base clock | 1130 MHz | 1410 MHz |
---|---|---|
Game clock | 1375 MHz | — |
Boost clock | 1560 MHz | 1665 MHz |
Mémoire clock | 1500 MHz 12 Gbps effective | 1750 MHz 14 Gbps effective |
Taille de la mémoire | 6 GB | 8 GB |
---|---|---|
Type de mémoire | GDDR6 | GDDR6 |
Bus mémoire | 192 bit | 256 bit |
Bande passante | 288.0 GB/s | 448.0 GB/s |
Shading units | 2304 | 4864 |
---|---|---|
TMUs | 144 | 152 |
ROPs | 64 | 80 |
Compute units | 36 | — |
Cache L2 | 3 MB | 4 MB |
Comptage SM | — | 38 |
Tensor cores | — | 152 |
RT cores | — | 38 |
Cache L1 | — | 128 KB (per SM) |
Taux de pixel | 99.84 GPixel/s | 133.2 GPixel/s |
---|---|---|
Taux de texture | 224.6 GTexel/s | 253.1 GTexel/s |
FP16 (half) performance | 14.38 TFLOPS (2:1) | 16.20 TFLOPS (1:1) |
FP32 (float) performance | 7.188 TFLOPS | 16.20 TFLOPS |
FP64 (double) performance | 449.3 GFLOPS (1:16) | 253.1 GFLOPS (1:64) |
Taille Slot | Dual-slot | Dual-slot |
---|---|---|
Longueur | 267 mm 10.5 inches | 242 mm 9.5 inches |
TDP | 150 W | 200 W |
Suggestions pour le PSU | 450 W | 550 W |
Sorties | 1x HDMI3x DisplayPort | 1x HDMI3x DisplayPort |
Connecteurs de puissance | 1x 8-pin | 1x 12-pin |
Taille | — | 112 mm 4.4 inches |
Board number | — | PG142 SKU 20 |
DirectX | 12 (12_1) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 2.1 | 3.0 |
Vulkan | 1.2 | 1.2 |
Shader model | 6.5 | 6.6 |
CUDA | — | 8.6 |