NVIDIA H100 CNX vs AMD Radeon PRO W7900
GPU · Specs Compare
Graphics Card
Release Date
Mar 2023
May 2023
Generation
Tesla Hopper
Radeon Pro Navi
Type
AI GPU
Professional
Bus Interface
PCIe 5.0 x16
PCIe 4.0 x16
Clock Speeds
Base Clock
690 MHz
1760 MHz
Boost Clock
1845 MHz
2495 MHz
Memory Clock
1593 MHz
2250 MHz
Memory
Memory Size
80GB
48GB
Memory Type
HBM2e
GDDR6
Memory Bus
5120bit
384bit
Bandwidth
2039GB/s
864.0GB/s
Render Config
SM Count
114
—
Shading Units
14592
6144
TMUs
456
384
ROPs
24
192
Tensor Cores
456
—
L1 Cache
256 KB (per SM)
256 KB per Array
L2 Cache
50 MB
6 MB
Compute Units
—
96
RT Cores
—
96
L3 Cache
—
96 MB
Theoretical Performance
Pixel Rate
44.28 GPixel/s
479.0 GPixel/s
Texture Rate
841.3 GTexel/s
958.1 GTexel/s
FP16 (half)
215.4 TFLOPS
122.6 TFLOPS
FP32 (float)
53.84 TFLOPS
61.32 TFLOPS
FP64 (double)
26.92 TFLOPS
1.916 TFLOPS
Graphics Processor
GPU Name
GH100
Navi 31
Architecture
Hopper
RDNA 3.0
Foundry
TSMC
TSMC
Process Size
4 nm
5 nm
Transistors
80 billion
57.7 billion
Die Size
814 mm²
529 mm²
GPU Variant
—
Navi 31 XTW
Board Design
TDP
350W
295W
Suggested PSU
750 W
600 W
Outputs
No outputs
3x DisplayPort 2.11x mini-DisplayPort 2.1
Power Connectors
8-pin EPS
2x 8-pin
Graphics Features
DirectX
N/A
12 Ultimate (12_2)
OpenGL
N/A
4.6
OpenCL
3.0
2.0
Vulkan
N/A
1.4
CUDA
9.0
—
Shader Model
N/A
6.9
▲ Specs Compare
Data sourced from official specifications and benchmark tests.