NVIDIA Tesla T10 Processor vs NVIDIA Tesla T4
Grafikprozessor
| GPU Name | GT200B | TU104 |
|---|---|---|
| Architektur | Tesla 2.0 | Turing |
| Hersteller | TSMC | TSMC |
| Prozessgröße | 55 nm | 12 nm |
| Transistoren | 1,400 million | 13,600 million |
| Chipgröße | 470 mm² | 545 mm² |
| GPU Variante | — | TU104-895-A1 |
Grafikkarte
| Veröffentlichungsdatum | Apr 9th, 2009 | Sep 13th, 2018 |
|---|---|---|
| Generation | Tesla | Tesla |
| Produktion | End-of-life | Active |
| Businterface | PCIe 2.0 x16 | PCIe 3.0 x16 |
Takt-Geschwindigkeiten
| GPU-Taktung | 610 MHz | — |
|---|---|---|
| Shader-Taktung | 1296 MHz | — |
| Speicher-Taktung | 800 MHz 1600 Mbps effective | 1250 MHz 10 Gbps effective |
| Basis-Taktfrequenz | — | 585 MHz |
| Boost-Takt | — | 1590 MHz |
Speicher
| Speichergröße | 4 GB | 16 GB |
|---|---|---|
| Speichertyp | GDDR3 | GDDR6 |
| Speicherbus | 512 bit | 256 bit |
| Bandbreite | 102.4 GB/s | 320.0 GB/s |
Render-Konfiguration
| Shading-Einheiten | 240 | 2560 |
|---|---|---|
| TMUs | 80 | 160 |
| ROPs | 32 | 64 |
| Anzahl SM | 30 | 40 |
| Cache L2 | 256 KB | 4 MB |
| Tensor Cores | — | 320 |
| RT-Kerne | — | 40 |
| Cache L1 | — | 64 KB (per SM) |
Theoretische Leistung
| Pixelrate | 19.52 GPixel/s | 101.8 GPixel/s |
|---|---|---|
| Texturrate | 48.80 GTexel/s | 254.4 GTexel/s |
| FP32 (float) Leistung | 622.1 GFLOPS | 8.141 TFLOPS |
| FP64 (double) Leistung | 77.76 GFLOPS (1:8) | 254.4 GFLOPS (1:32) |
| FP16 (half) Leistung | — | 65.13 TFLOPS (8:1) |
Boarddesign
| Steckplatzbreite | Dual-slot | Single-slot |
|---|---|---|
| Länge | 267 mm 10.5 inches | 168 mm 6.6 inches |
| TDP | 188 W | 70 W |
| Vorgeschlagenes Netzteil | 450 W | 250 W |
| Ausgabe | No outputs | No outputs |
| Stromverbindung | — | None |
| Board-Nummer | — | PG183 SKU 200 |
Grafik-Features
| DirectX | 11.1 (10_0) | 12 Ultimate (12_2) |
|---|---|---|
| OpenGL | 3.3 | 4.6 |
| OpenCL | 1.1 | 3.0 |
| Vulkan | — | 1.2 |
| CUDA | 1.3 | 7.5 |
| Shader-Modell | 4.0 | 6.6 |
