AMD Radeon Instinct MI308X vs NVIDIA H100 CNX
GPU Specs Compare
Graphics Card
Release Date
Dec 2023
Mar 2023
Generation
Radeon Instinct
Tesla Hopper
Type
Professional
AI GPU
Bus Interface
PCIe 5.0 x16
PCIe 5.0 x16
Clock Speeds
Base Clock
1000 MHz
690 MHz
Boost Clock
2100 MHz
1845 MHz
Memory Clock
1300 MHz
1593 MHz
Memory
Memory Size
192GB
80GB
Memory Type
HBM3
HBM2e
Memory Bus
8192bit
5120bit
Bandwidth
5.32TB/s
2039GB/s
Render Config
Compute Units
304
-
Shading Units
19456
14592
TMUs
1216
456
ROPs
0
24
Tensor Cores
1216
456
L1 Cache
16 KB (per CU)
256 KB (per SM)
L2 Cache
16 MB
50 MB
L3 Cache
256 MB
-
SM Count
-
114
Theoretical Performance
Pixel Rate
0 MPixel/s
44.28 GPixel/s
Texture Rate
2553.6 GTexel/s
841.3 GTexel/s
FP16 (half)
81.72 TFLOPS
215.4 TFLOPS
FP32 (float)
81.72 TFLOPS
53.84 TFLOPS
FP64 (double)
40.86 TFLOPS
26.92 TFLOPS
Graphics Processor
GPU Name
Aqua Vanjaram
GH100
Architecture
CDNA 3.0
Hopper
Foundry
TSMC
TSMC
Process Size
5 nm
4 nm
Transistors
153 billion
80 billion
Die Size
1017 mm²
814 mm²
Board Design
TDP
750W
350W
Suggested PSU
1150 W
750 W
Outputs
No outputs
No outputs
Power Connectors
None
8-pin EPS
Graphics Features
DirectX
N/A
N/A
OpenGL
N/A
N/A
OpenCL
3.0
3.0
Vulkan
N/A
N/A
Shader Model
N/A
N/A
CUDA
-
9.0