Nom du GPU | TU117 | GA106 |
---|---|---|
GPU variant | N18S-G5 | GA106-140-A1 |
Architecture | Turing | Ampere |
Fonderie | TSMC | Samsung |
Taille du processus | 12 nm | 8 nm |
Transistors | 4,700 million | 13,250 million |
Taille des matrices | 200 mm² | 276 mm² |
Date de publication | Aug 15th, 2020 | — |
---|---|---|
Génération | GeForce MX (4xx) | — |
Production | Active | — |
Bus interface | PCIe 4.0 x4 | — |
Base clock | 720 MHz | 1545 MHz |
---|---|---|
Boost clock | 930 MHz | 1740 MHz |
Mémoire clock | 1250 MHz 10 Gbps effective | 1750 MHz 14 Gbps effective |
Taille de la mémoire | 2 GB | 4 GB |
---|---|---|
Type de mémoire | GDDR6 | GDDR6 |
Bus mémoire | 64 bit | 128 bit |
Bande passante | 80.00 GB/s | 224.0 GB/s |
Shading units | 896 | 2304 |
---|---|---|
TMUs | 56 | 72 |
ROPs | 32 | 40 |
Comptage SM | 14 | 18 |
Cache L1 | 64 KB (per SM) | 128 KB (per SM) |
Cache L2 | 512 KB | 2 MB |
Tensor cores | — | 72 |
RT cores | — | 18 |
Taux de pixel | 29.76 GPixel/s | 69.60 GPixel/s |
---|---|---|
Taux de texture | 52.08 GTexel/s | 125.3 GTexel/s |
FP16 (half) performance | 3.333 TFLOPS (2:1) | 8.018 TFLOPS (1:1) |
FP32 (float) performance | 1.667 TFLOPS | 8.018 TFLOPS |
FP64 (double) performance | 52.08 GFLOPS (1:32) | 125.3 GFLOPS (1:64) |
TDP | 25 W | 90 W |
---|---|---|
Sorties | No outputs | 1x HDMI3x DisplayPort |
Connecteurs de puissance | None | 1x 6-pin |
Taille Slot | — | Dual-slot |
Longueur | — | 242 mm 9.5 inches |
Taille | — | 112 mm 4.4 inches |
Suggestions pour le PSU | — | 250 W |
DirectX | 12 (12_1) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 3.0 | 3.0 |
Vulkan | 1.2 | 1.2 |
CUDA | 7.5 | 8.6 |
Shader model | 6.6 | 6.6 |
Date de publication | — | 2022 |
---|---|---|
gpu.details.availability | — | 2022 |
Génération | — | GeForce 30 |
Prédécesseur | — | GeForce 20 |
Production | — | Unreleased |
Bus interface | — | PCIe 4.0 x16 |