GPU Name | NV30 | GV100 |
---|---|---|
GPU Variante | NV30GL | — |
Architektur | Rankine | Volta |
Hersteller | TSMC | TSMC |
Prozessgröße | 130 nm | 12 nm |
Transistoren | 125 million | 21,100 million |
Chipgröße | 199 mm² | 815 mm² |
Veröffentlichungsdatum | Jan 21st, 2003 | Mar 27th, 2018 |
---|---|---|
Generation | Quadro FX | Tesla |
Produktion | End-of-life | End-of-life |
Businterface | AGP 8x | PCIe 3.0 x16 |
Reviews | 26 in our database | — |
GPU-Taktung | 400 MHz | — |
---|---|---|
Speicher-Taktung | 400 MHz 800 Mbps effective | 876 MHz 1752 Mbps effective |
Basis-Taktfrequenz | — | 1297 MHz |
Boost-Takt | — | 1530 MHz |
Speichergröße | 128 MB | 32 GB |
---|---|---|
Speichertyp | DDR2 | HBM2 |
Speicherbus | 128 bit | 4096 bit |
Bandbreite | 12.80 GB/s | 897.0 GB/s |
Pixelshader | 4 | — |
---|---|---|
Vertex-Shader | 2 | — |
TMUs | 8 | 320 |
ROPs | 4 | 128 |
Shading-Einheiten | — | 5120 |
Anzahl SM | — | 80 |
Tensor Cores | — | 640 |
Cache L1 | — | 128 KB (per SM) |
Cache L2 | — | 6 MB |
Pixelrate | 1.600 GPixel/s | 195.8 GPixel/s |
---|---|---|
Vertex-Rate | 200.0 MVertices/s | — |
Texturrate | 3.200 GTexel/s | 489.6 GTexel/s |
FP16 (half) Leistung | — | 31.33 TFLOPS (2:1) |
FP32 (float) Leistung | — | 15.67 TFLOPS |
FP64 (double) Leistung | — | 7.834 TFLOPS (1:2) |
Steckplatzbreite | Single-slot | Dual-slot |
---|---|---|
Länge | 229 mm 9 inches | — |
Breite | 111 mm 4.4 inches | — |
TDP | unknown | 250 W |
Vorgeschlagenes Netzteil | 200 W | 600 W |
Ausgabe | 2x DVI1x S-Video | No outputs |
Stromverbindung | 1x Molex | None |
DirectX | 9.0a | 12 (12_1) |
---|---|---|
OpenGL | 1.5 (2.1) | 4.6 |
OpenCL | — | 3.0 |
Vulkan | — | 1.2 |
Pixelshader | 2.0a | — |
Vertex-Shader | 2.0a | — |
CUDA | — | 7.0 |
Shader-Modell | — | 6.6 |