GPU Name | GA102 | GP102 |
---|---|---|
GPU Variante | GA102-350-A1 | — |
Architektur | Ampere | Pascal |
Hersteller | Samsung | TSMC |
Prozessgröße | 8 nm | 16 nm |
Transistoren | 28,300 million | 11,800 million |
Chipgröße | 628 mm² | 471 mm² |
Veröffentlichungsdatum | Unknown | Sep 13th, 2016 |
---|---|---|
gpu.details.availability | 2022 | — |
Generation | GeForce 30 | Tesla |
Vorgänger | GeForce 20 | — |
Produktion | Unreleased | End-of-life |
Originalpreis | 1,699 USD | 5,699 USD |
Businterface | PCIe 4.0 x16 | PCIe 3.0 x16 |
Reviews | 55 in our database | — |
Basis-Taktfrequenz | 1560 MHz | 1303 MHz |
---|---|---|
Boost-Takt | 1860 MHz | 1531 MHz |
Speicher-Taktung | 1325 MHz 21.2 Gbps effective | 1808 MHz 14.5 Gbps effective |
Speichergröße | 24 GB | 24 GB |
---|---|---|
Speichertyp | GDDR6X | GDDR5X |
Speicherbus | 384 bit | 384 bit |
Bandbreite | 1,018 GB/s | 694.3 GB/s |
Shading-Einheiten | 10752 | 3840 |
---|---|---|
TMUs | 336 | 240 |
ROPs | 112 | 96 |
Anzahl SM | 84 | 30 |
Tensor Cores | 336 | — |
RT-Kerne | 84 | — |
Cache L1 | 128 KB (per SM) | 48 KB (per SM) |
Cache L2 | 6 MB | 3 MB |
Pixelrate | 208.3 GPixel/s | 147.0 GPixel/s |
---|---|---|
Texturrate | 625.0 GTexel/s | 367.4 GTexel/s |
FP16 (half) Leistung | 40.00 TFLOPS (1:1) | 183.7 GFLOPS (1:64) |
FP32 (float) Leistung | 40.00 TFLOPS | 11.76 TFLOPS |
FP64 (double) Leistung | 625.0 GFLOPS (1:64) | 367.4 GFLOPS (1:32) |
Steckplatzbreite | Triple-slot | Dual-slot |
---|---|---|
Länge | 336 mm 13.2 inches | 267 mm 10.5 inches |
Breite | 140 mm 5.5 inches | 111 mm 4.4 inches |
Höhe | 61 mm 2.4 inches | — |
TDP | 450 W | 250 W |
Vorgeschlagenes Netzteil | 850 W | 600 W |
Ausgabe | 1x HDMI3x DisplayPort | No outputs |
Stromverbindung | 1x 12-pin | 1x 6-pin + 1x 8-pin |
Board-Nummer | PG132 SKU | PG610 SKU 200 |
DirectX | 12 Ultimate (12_2) | 12 (12_1) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 3.0 | 3.0 |
Vulkan | 1.2 | 1.2 |
CUDA | 8.6 | 6.1 |
Shader-Modell | 6.6 | 6.4 |