AMD Radeon Instinct MI210 vs NVIDIA GeForce GTX 1050 3 GB
Processeur graphique
| Nom du GPU | Aldebaran | GP107 |
|---|---|---|
| GPU variant | Aldebaran | GP107-301-K1-A1 |
| Architecture | CDNA 2.0 | Pascal |
| Fonderie | TSMC | Samsung |
| Taille du processus | 6 nm | 14 nm |
| Transistors | 58,200 million | 3,300 million |
| Taille des matrices | unknown | 132 mm² |
Carte graphique
| Date de publication | Dec 2021 | May 21st, 2018 |
|---|---|---|
| Génération | Radeon Instinct | GeForce 10 |
| Production | Active | End-of-life |
| Bus interface | PCIe 4.0 x16 | PCIe 3.0 x16 |
| Prédécesseur | — | GeForce 900 |
| Successeur | — | GeForce 20 |
| Critiques | — | 58 in our database |
Vitesse clock
| Base clock | 1000 MHz | 1392 MHz |
|---|---|---|
| Boost clock | 1700 MHz | 1518 MHz |
| Mémoire clock | 1600 MHz 3.2 Gbps effective | 1752 MHz 7 Gbps effective |
Mémoire
| Taille de la mémoire | 64 GB | 3 GB |
|---|---|---|
| Type de mémoire | HBM2e | GDDR5 |
| Bus mémoire | 4096 bit | 96 bit |
| Bande passante | 1,638 GB/s | 84.10 GB/s |
Render Config
| Shading units | 6656 | 768 |
|---|---|---|
| TMUs | 416 | 48 |
| ROPs | 0 | 24 |
| Compute units | 104 | — |
| gpu.details.mcm | 1 | — |
| Cache L1 | 16 KB (per CU) | 48 KB (per SM) |
| Cache L2 | 16 MB | 768 KB |
| Comptage SM | — | 6 |
Performance théorique
| Taux de pixel | 0 MPixel/s | 36.43 GPixel/s |
|---|---|---|
| Taux de texture | 707.2 GTexel/s | 72.86 GTexel/s |
| FP16 (half) performance | 181.0 TFLOPS (8:1) | 36.43 GFLOPS (1:64) |
| FP32 (float) performance | 22.63 TFLOPS | 2.332 TFLOPS |
| FP64 (double) performance | 22.63 TFLOPS (1:1) | 72.86 GFLOPS (1:32) |
Board Design
| Taille Slot | Dual-slot | Dual-slot |
|---|---|---|
| Longueur | 267 mm 10.5 inches | 145 mm 5.7 inches |
| Taille | 111 mm 4.4 inches | 111 mm 4.4 inches |
| TDP | 300 W | 75 W |
| Suggestions pour le PSU | 700 W | 250 W |
| Sorties | No outputs | 1x DVI1x HDMI1x DisplayPort |
| Connecteurs de puissance | 2x 8-pin | None |
| Board number | — | PG210 |
Caractéristiques graphiques
| DirectX | — | 12 (12_1) |
|---|---|---|
| OpenGL | — | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | — | 1.2 |
| Shader model | — | 6.4 |
| CUDA | — | 6.1 |
