GPU model | Gemini Lake GT1 | GA102 |
---|---|---|
Architecture | Generation 9.5 | Ampere |
Foundry | Intel | Samsung |
Manufacturing process | 14 nm | 8 nm |
Number of transistors | unknown | 28,300 million |
Die size | unknown | 628 mm² |
GPU variant | — | GA102-890-A1 |
Release date | Dec 11th, 2017 | — |
---|---|---|
Generation | HD Graphics-T (Goldmont Plus) | — |
Production | Active | — |
Interface | Ring Bus | — |
Reviews | 1 in our database | — |
Base clock | 200 MHz | 885 MHz |
---|---|---|
Boost clock | 650 MHz | 1695 MHz |
Memory clock | System Shared | 1563 MHz 12.5 Gbps effective |
Memory size | System Shared | 24 GB |
---|---|---|
Memory type | System Shared | GDDR6 |
Memory bus width | System Shared | 384 bit |
Bandwidth | System Dependent | 600.2 GB/s |
Shading units | 96 | 9216 |
---|---|---|
TMUs | 12 | 288 |
ROPs | 2 | 96 |
Execution units | 12 | — |
SM count | — | 72 |
Tensor cores | — | 288 |
RT cores | — | 72 |
Cache L1 | — | 128 KB (per SM) |
Cache L2 | — | 6 MB |
Pixel rate | 1.300 GPixel/s | 162.7 GPixel/s |
---|---|---|
Texture rate | 7.800 GTexel/s | 488.2 GTexel/s |
FP16 (half) performance | 249.6 GFLOPS (2:1) | 31.24 TFLOPS (1:1) |
FP32 (float) performance | 124.8 GFLOPS | 31.24 TFLOPS |
FP64 (double) performance | 31.20 GFLOPS (1:4) | 976.3 GFLOPS (1:32) |
Slot width | IGP | Single-slot |
---|---|---|
TDP | 5 W | 150 W |
Outputs | No outputs | No outputs |
Length | — | 267 mm 10.5 inches |
Width | — | 112 mm 4.4 inches |
Suggested PSU | — | 450 W |
Power connectors | — | 8-pin EPS |
Board number | — | PG133 SKU 215 |
DirectX | 12 (12_1) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 3.0 | 3.0 |
Vulkan | 1.2 | 1.2 |
Shader model | 6.4 | 6.6 |
CUDA | — | 8.6 |
Release date | — | Apr 12th, 2021 |
---|---|---|
gpu.details.availability | — | 2021 |
Generation | — | Tesla |
Production | — | Active |
Interface | — | PCIe 4.0 x16 |