GPU Wiki

Compare GPUs

Select up to four GPU Wiki entries to compare matching specifications row by row. The comparison uses the site database and keeps generation-specific specs visibly separate.

Popular comparisons

Add a third or fourth GPU
Spec CMP 90HX
Overview
Product name NVIDIA CMP 90HX
Architecture Ampere
Generation / family Mining GPUs
GPU codename GA102
Launch date 2021-02-18
Market segment Desktop in Mining GPUs
Board type / entry type Reference GPU
GPU
Process node Samsung 8 nm
Die size 628 mm²
Transistor count 28.3 billion
GPU chip GA102
SM / multiprocessor count 50
CUDA Cores 6400
Tensor Cores 200
RT Cores 50
ROPs 80
TMUs 200
Clocks
Base / Core clock 1,500 MHz
Boost clock 1,710 MHz
Memory clock 1,188 MHz
Memory
Memory size 10 GB
Memory type GDDR6X
Memory bus width 320-bit
Memory bandwidth 760.3 GB/s
Bus / Interface
Bus interface PCIe 1.0 x4
AGP / PCI / PCIe distinction PCIe
Power
TDP / TBP / board power 320 W
Recommended PSU 700 W
Power connectors 2x 8-pin
Slot width Dual-slot, 285 mm length
Features
DirectX support DirectX 12.2
OpenGL support OpenGL 4.6
Vulkan support Vulkan 1.4
CUDA support 8.6
Shader Model 6.8
Ray tracing support Hardware RT units present
Performance / Benchmarks
FP32 throughput 21.89 TFLOPS
FP16 throughput 21.89 TFLOPS
Muted cells indicate unavailable, not applicable, or not directly comparable specifications.

Frame rates compare runs of the same game at the same quality preset on one test system. DLAA counts as native resolution, because it applies anti-aliasing at full render resolution rather than upscaling; DLSS and FSR upscaling are never counted as native. Where a capture named ray tracing without a readable on or off value, it is read as on.