NVIDIA GRID A100A vs NVIDIA Quadro K4000
Grafikprozessor
| GPU Name | GA100 | GK106 |
|---|---|---|
| Architektur | Ampere | Kepler |
| Hersteller | TSMC | TSMC |
| Prozessgröße | 7 nm | 28 nm |
| Transistoren | 54,200 million | 2,540 million |
| Chipgröße | 826 mm² | 221 mm² |
Grafikkarte
| Veröffentlichungsdatum | May 14th, 2020 | Mar 1st, 2013 |
|---|---|---|
| Generation | GRID | Quadro |
| Produktion | Active | End-of-life |
| Businterface | PCIe 4.0 x16 | PCIe 2.0 x16 |
| Originalpreis | — | 1,269 USD |
| Reviews | — | 26 in our database |
Takt-Geschwindigkeiten
| Basis-Taktfrequenz | 900 MHz | — |
|---|---|---|
| Boost-Takt | 1005 MHz | — |
| Speicher-Taktung | 1215 MHz 2.4 Gbps effective | 1404 MHz 5.6 Gbps effective |
| GPU-Taktung | — | 810 MHz |
Speicher
| Speichergröße | 48 GB | 3 GB |
|---|---|---|
| Speichertyp | HBM2E | GDDR5 |
| Speicherbus | 6144 bit | 192 bit |
| Bandbreite | 1,866 GB/s | 134.8 GB/s |
Render-Konfiguration
| Shading-Einheiten | 6912 | 768 |
|---|---|---|
| TMUs | 432 | 64 |
| ROPs | 192 | 24 |
| Anzahl SM | 108 | — |
| Tensor Cores | 432 | — |
| Cache L1 | 192 KB (per SM) | 16 KB (per SMX) |
| Cache L2 | 48 MB | 384 KB |
| Anzahl SMX | — | 4 |
Theoretische Leistung
| Pixelrate | 193.0 GPixel/s | 12.96 GPixel/s |
|---|---|---|
| Texturrate | 434.2 GTexel/s | 51.84 GTexel/s |
| FP16 (half) Leistung | 55.57 TFLOPS (4:1) | — |
| FP32 (float) Leistung | 13.89 TFLOPS | 1,244 GFLOPS |
| FP64 (double) Leistung | 6.947 TFLOPS (1:2) | 51.84 GFLOPS (1:24) |
Boarddesign
| Steckplatzbreite | IGP | Single-slot |
|---|---|---|
| TDP | 400 W | 80 W |
| Vorgeschlagenes Netzteil | 800 W | 250 W |
| Ausgabe | No outputs | 1x DVI2x DisplayPort |
| Stromverbindung | None | 1x 6-pin |
| Länge | — | 241 mm 9.5 inches |
| Breite | — | 111 mm 4.4 inches |
| Board-Nummer | — | P2030 |
Grafik-Features
| DirectX | — | 12 (11_0) |
|---|---|---|
| OpenGL | — | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | — | 1.1 |
| CUDA | 8.0 | 3.0 |
| Shader-Modell | — | 5.1 |
