GPU Name | Aldebaran | GA100 |
---|---|---|
GPU Variante | Aldebaran | — |
Architektur | CDNA 2.0 | Ampere |
Hersteller | TSMC | TSMC |
Prozessgröße | 6 nm | 7 nm |
Transistoren | 58,200 million | 54,200 million |
Chipgröße | unknown | 826 mm² |
Veröffentlichungsdatum | Dec 2021 | Jun 22nd, 2020 |
---|---|---|
Generation | Radeon Instinct | Tesla |
Produktion | Active | Active |
Businterface | PCIe 4.0 x16 | PCIe 4.0 x16 |
Basis-Taktfrequenz | 1000 MHz | 765 MHz |
---|---|---|
Boost-Takt | 1700 MHz | 1410 MHz |
Speicher-Taktung | 1600 MHz 3.2 Gbps effective | 1215 MHz 2.4 Gbps effective |
GPU-Taktung | — | 1410 MHz |
Speichergröße | 64 GB | 40 GB |
---|---|---|
Speichertyp | HBM2e | HBM2E |
Speicherbus | 4096 bit | 5120 bit |
Bandbreite | 1,638 GB/s | 1,555 GB/s |
Shading-Einheiten | 6656 | 6912 |
---|---|---|
TMUs | 416 | 432 |
ROPs | 0 | 160 |
Compute Units | 104 | — |
gpu.details.mcm | 1 | — |
Cache L1 | 16 KB (per CU) | 192 KB (per SM) |
Cache L2 | 16 MB | 40 MB |
Anzahl SM | — | 108 |
Tensor Cores | — | 432 |
Pixelrate | 0 MPixel/s | 225.6 GPixel/s |
---|---|---|
Texturrate | 707.2 GTexel/s | 609.1 GTexel/s |
FP16 (half) Leistung | 181.0 TFLOPS (8:1) | 77.97 TFLOPS (4:1) |
FP32 (float) Leistung | 22.63 TFLOPS | 19.49 TFLOPS |
FP64 (double) Leistung | 22.63 TFLOPS (1:1) | 9.746 TFLOPS (1:2) |
Steckplatzbreite | Dual-slot | Dual-slot |
---|---|---|
Länge | 267 mm 10.5 inches | 267 mm 10.5 inches |
Breite | 111 mm 4.4 inches | — |
TDP | 300 W | 250 W |
Vorgeschlagenes Netzteil | 700 W | 600 W |
Ausgabe | No outputs | No outputs |
Stromverbindung | 2x 8-pin | 8-pin EPS |
Board-Nummer | — | P1001 SKU 200 |
DirectX | — | — |
---|---|---|
OpenGL | — | — |
OpenCL | 3.0 | 3.0 |
Vulkan | — | — |
Shader-Modell | — | — |
CUDA | — | 8.0 |