NVIDIA A100 PCIe 80 GB vs NVIDIA GeForce RTX 3080 Ti
Processeur graphique
| Nom du GPU | GA100 | GA102 |
|---|---|---|
| Architecture | Ampere | Ampere |
| Fonderie | TSMC | TSMC |
| Taille du processus | 7 nm | 7 nm |
| Transistors | 54,200 million | 40,000 million |
| Taille des matrices | 826 mm² | 627 mm² |
| GPU variant | — | GA102-300-A1 |
Carte graphique
| Date de publication | Jun 28th, 2021 | 2020 |
|---|---|---|
| Génération | Tesla | GeForce 30 |
| Production | Active | Unreleased |
| Bus interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
| Prix de lancement | — | 1,399 USD |
Vitesse clock
| Base clock | 1065 MHz | 1410 MHz |
|---|---|---|
| Boost clock | 1410 MHz | 1740 MHz |
| Mémoire clock | 1593 MHz 3.2 Gbps effective | 2375 MHz 19000 MHz effective |
Mémoire
| Taille de la mémoire | 80 GB | 12 GB |
|---|---|---|
| Type de mémoire | HBM2e | GDDR6X |
| Bus mémoire | 5120 bit | 384 bit |
| Bande passante | 2,039 GB/s | 912.0 GB/s |
Render Config
| Shading units | 6912 | 5120 |
|---|---|---|
| TMUs | 432 | 320 |
| ROPs | 160 | 96 |
| Comptage SM | 108 | 80 |
| Tensor cores | 432 | 640 |
| Cache L1 | 192 KB (per SM) | 192 KB (per SM) |
| Cache L2 | 80 MB | 6 MB |
| RT cores | — | 160 |
Performance théorique
| Taux de pixel | 225.6 GPixel/s | 167.0 GPixel/s |
|---|---|---|
| Taux de texture | 609.1 GTexel/s | 556.8 GTexel/s |
| FP16 (half) performance | 77.97 TFLOPS (4:1) | 35.64 TFLOPS (2:1) |
| FP32 (float) performance | 19.49 TFLOPS | 17.82 TFLOPS |
| FP64 (double) performance | 9.746 TFLOPS (1:2) | 556.8 GFLOPS (1:32) |
Board Design
| Taille Slot | Dual-slot | Dual-slot |
|---|---|---|
| Longueur | 267 mm 10.5 inches | 267 mm 10.5 inches |
| TDP | 250 W | 320 W |
| Suggestions pour le PSU | 600 W | 700 W |
| Sorties | No outputs | 1x HDMI3x DisplayPort1x USB Type-C |
| Connecteurs de puissance | 8-pin EPS | 1x 12-pin |
| Board number | P1001 SKU 200 | PG132 SKU 20 |
Caractéristiques graphiques
| DirectX | — | 12 Ultimate (12_2) |
|---|---|---|
| OpenGL | — | 4.6 |
| OpenCL | 3.0 | 2.0 |
| Vulkan | — | 1.2.140 |
| CUDA | 8.0 | 8.5 |
| Shader model | — | 6.5 |
