Nom du GPU | Oberon | GA102 |
---|---|---|
Architecture | RDNA 2.0 | Ampere |
Fonderie | TSMC | Samsung |
Taille du processus | 7 nm | 8 nm |
Transistors | unknown | 28,300 million |
Taille des matrices | 308 mm² | 628 mm² |
GPU variant | CXD90044GB | GA102-200-KD-A1 |
Date de publication | Nov 12th, 2020 | Sep 1st, 2020 |
---|---|---|
Génération | Console GPU | GeForce 30 |
Production | Active | Active |
Prix de lancement | 499 USD | 699 USD |
Bus interface | IGP | PCIe 4.0 x16 |
gpu.details.availability | — | Sep 17th, 2020 |
Prédécesseur | — | GeForce 20 |
Prix courant | — | Amazon / Newegg |
Critiques | — | 99 in our database |
GPU clock | 2233 MHz | — |
---|---|---|
Mémoire clock | 1750 MHz 14 Gbps effective | 1188 MHz 19 Gbps effective |
Base clock | — | 1440 MHz |
Boost clock | — | 1710 MHz |
Taille de la mémoire | 16 GB | 10 GB |
---|---|---|
Type de mémoire | GDDR6 | GDDR6X |
Bus mémoire | 256 bit | 320 bit |
Bande passante | 448.0 GB/s | 760.3 GB/s |
Shading units | 2304 | 8704 |
---|---|---|
TMUs | 144 | 272 |
ROPs | 64 | 96 |
Compute units | 36 | — |
Cache L2 | 4 MB | 5 MB |
Comptage SM | — | 68 |
Tensor cores | — | 272 |
RT cores | — | 68 |
Cache L1 | — | 128 KB (per SM) |
Taux de pixel | 142.9 GPixel/s | 164.2 GPixel/s |
---|---|---|
Taux de texture | 321.6 GTexel/s | 465.1 GTexel/s |
FP16 (half) performance | 20.58 TFLOPS (2:1) | 29.77 TFLOPS (1:1) |
FP32 (float) performance | 10.29 TFLOPS | 29.77 TFLOPS |
FP64 (double) performance | 643.1 GFLOPS (1:16) | 465.1 GFLOPS (1:64) |
Taille Slot | IGP | Dual-slot |
---|---|---|
Longueur | 390 mm 15.4 inches | 285 mm 11.2 inches |
Taille | 260 mm 10.2 inches | 112 mm 4.4 inches |
Taille | 104 mm 4.1 inches | 40 mm 1.6 inches |
TDP | 180 W | 320 W |
Suggestions pour le PSU | 450 W | 700 W |
Sorties | No outputs | 1x HDMI3x DisplayPort |
Connecteurs de puissance | None | 1x 12-pin |
Poids | 4.5 kg (9.9 lbs) | — |
Board number | — | PG132 SKU 30 |
DirectX | — | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 1.2 | 3.0 |
Vulkan | 1.2 | 1.2 |
Shader model | 6.5 | 6.6 |
CUDA | — | 8.6 |