GPU Name | Kalindi | TU106 |
---|---|---|
Architektur | GCN 2.0 | Turing |
Hersteller | TSMC | TSMC |
Prozessgröße | 28 nm | 12 nm |
Transistoren | 1,178 million | 10,800 million |
Chipgröße | 110 mm² | 445 mm² |
GPU Variante | — | N19E-Q1-KD-A1 |
Veröffentlichungsdatum | Apr 23rd, 2013 | — |
---|---|---|
Generation | Kabini (HD 8000 Mobile) | — |
Produktion | End-of-life | — |
Businterface | IGP | — |
Vorgänger | Trinity | — |
Nachfolger | Kaveri | — |
GPU-Taktung | 450 MHz | — |
---|---|---|
Speicher-Taktung | System Shared | 1750 MHz 14 Gbps effective |
Basis-Taktfrequenz | — | 945 MHz |
Boost-Takt | — | 1380 MHz |
Speichergröße | System Shared | 6 GB |
---|---|---|
Speichertyp | System Shared | GDDR6 |
Speicherbus | System Shared | 192 bit |
Bandbreite | System Dependent | 336.0 GB/s |
Shading-Einheiten | 128 | 1920 |
---|---|---|
TMUs | 8 | 120 |
ROPs | 4 | 64 |
Compute Units | 2 | — |
Anzahl SM | — | 30 |
Tensor Cores | — | 240 |
RT-Kerne | — | 30 |
Cache L1 | — | 64 KB (per SM) |
Cache L2 | — | 4 MB |
Pixelrate | 1.800 GPixel/s | 88.32 GPixel/s |
---|---|---|
Texturrate | 3.600 GTexel/s | 165.6 GTexel/s |
FP32 (float) Leistung | 115.2 GFLOPS | 5.299 TFLOPS |
FP64 (double) Leistung | 7.200 GFLOPS (1:16) | 165.6 GFLOPS (1:32) |
FP16 (half) Leistung | — | 10.60 TFLOPS (2:1) |
Steckplatzbreite | IGP | IGP |
---|---|---|
TDP | 15 W | 80 W |
Ausgabe | No outputs | No outputs |
Stromverbindung | — | None |
DirectX | 12 (12_0) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 2.0 | 3.0 |
Vulkan | 1.2 | 1.2 |
Shader-Modell | 6.3 | 6.6 |
CUDA | — | 7.5 |
Veröffentlichungsdatum | — | May 27th, 2019 |
---|---|---|
Generation | — | Quadro Mobile (Tx000) |
Produktion | — | Active |
Businterface | — | PCIe 3.0 x16 |