Report an Error

NVIDIA GH100

Default
GH100
Block Diagram
Block Diagram
SM Diagram
SM Diagram
NVIDIA's GH100 GPU uses the Hopper architecture and is made using a 5 nm production process at TSMC. With a die size of 814 mm² and a transistor count of 80,000 million it is a very big chip. GH100 does not support DirectX. For GPU compute applications, OpenCL version 3.0 and CUDA 9.0 can be used. It features 18432 shading units, 576 texture mapping units and 24 ROPs. Also included are 576 tensor cores which help improve the speed of machine learning applications.
Further reading: Hopper Architecture Whitepaper

Graphics Processor

Released
Mar 21st, 2023
GPU Name
GH100
Codename
NV180
Architecture
Hopper
Foundry
TSMC
Process Type
4N FinFET
Process Size
5 nm
Transistors
80,000 million
Density
98.3M / mm²
Die Size
814 mm²

Graphics Features

DirectX
N/A
OpenGL
N/A
OpenCL
3.0
Vulkan
N/A
CUDA
9.0
Shader Model
N/A
Tensor Cores
4th Gen
PureVideo HD
VP11
VDPAU
Feature Set K

Render Config

Shading Units
18432
TMUs
576
ROPs
24
SM Count
132
FP16 Units
73,728
FP64 Units
9216
INT32 Units
9216
Tensor Cores
576
Tex L1 Cache
64 KB per SM
L1 Cache
256 KB per SM
L2 Cache
61440 KB
Max. TDP
700 W
SFUs
2304
TPCs
72
GPCs
8

All Hopper GPUs

  • NVIDIA GH100

NVIDIA GPU Architecture History

Graphics cards using the NVIDIA GH100 GPU

Name Chip Memory Shaders TMUs ROPs Base Clock Boost Clock Memory Clock
80 GB 14592 456 24 1095 MHz 1755 MHz 1593 MHz
80 GB 16896 528 24 1590 MHz 1980 MHz 1313 MHz
96 GB 16896 528 24 1665 MHz 1837 MHz 1313 MHz
80 GB 16896 528 24 1095 MHz 1755 MHz 1313 MHz
80 GB 14592 456 24 690 MHz 1845 MHz 1593 MHz
96 GB 16896 528 24 1665 MHz 1837 MHz 1313 MHz
64 GB 16896 528 24 1665 MHz 1980 MHz 1313 MHz
80 GB 14592 456 24 1095 MHz 1755 MHz 1593 MHz

GH100 GPU Notes

Tensor Cores: 4th Gen
NVENC: No Support
NVDEC: No Support
PureVideo HD: VP11
VDPAU: Feature Set K
Jun 2nd, 2024 19:02 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts