NVIDIA GA102
NVIDIA's GA102 GPU uses the Ampere architecture and is made using a 8 nm production process at Samsung. With a die size of 628 mm² and a transistor count of 28,300 million it is a very big chip. GA102 supports DirectX 12 Ultimate (Feature Level 12_2). For GPU compute applications, OpenCL version 2.0 and CUDA 8.6 can be used. Additionally, the DirectX 12 Ultimate capability guarantees support for hardware-raytracing, variable-rate shading and more, in upcoming video games. It features 10752 shading units, 336 texture mapping units and 112 ROPs. Also included are 336 tensor cores which help improve the speed of machine learning applications. The GPU also contains 84 raytracing acceleration cores.
Further reading:
Ampere Architecture Whitepaper

Graphics Processor
- GPU Name
- GA102
- Codename
- NV172
- Architecture
- Ampere
- Foundry
- Samsung
- Process Size
- 8 nm
- Transistors
- 28,300 million
- Density
- 45.1M / mm²
- Die Size
- 628 mm²
- Released
- Sep 1st, 2020
Graphics Features
- DirectX
- 12 Ultimate (12_2)
- OpenGL
- 4.6
- OpenCL
- 2.0
- Vulkan
- 1.2
- CUDA
- 8.6
- Shader Model
- 6.5
- PureVideo HD
- VP11
- VDPAU
- Feature Set k
Render Config
- Shading Units
- 10752
- TMUs
- 336
- ROPs
- 112
- SM Count
- 84
- FP16 Units
- 10752
- FP64 Units
- 168
- INT32 Units
- 5376
- Tensor Cores
- 336
- RT Cores
- 84
- SFUs
- 1344
- TPCs
- 42
- GPCs
- 7
- Tex L1 Cache
- 64 KB per SM
- L1 Cache
- 128 KB per SM
- L2 Cache
- 6144 KB
- Register File
- 21504 KB
- Max. TDP
- 350 W
All Ampere GPUs
- NVIDIA GA100
- NVIDIA GA102
- NVIDIA GA103
- NVIDIA GA104
- NVIDIA GA106
- NVIDIA GA107
NVIDIA GPU Architecture History
- 1998-2000 Fahrenheit
- 1999-2005 Celsius
- 2001-2003 Kelvin
- 2003-2005 Rankine
- 2003-2013 Curie
- 2006-2010 Tesla
- 2007-2013 Tesla 2.0
- 2010-2016 Fermi
- 2010-2013 VLIW Vec4
- 2010-2016 Fermi 2.0
- 2012-2018 Kepler
- 2013-2015 Kepler 2.0
- 2014-2017 Maxwell
- 2014-2019 Maxwell 2.0
- 2016-2021 Pascal
- 2017-2020 Volta
- 2018-2021 Turing
- 2020-2021 Ampere
Graphics cards using the NVIDIA GA102 GPU
Name | Chip | Memory | Shaders | TMUs | ROPs | Base Clock | Boost Clock | Memory Clock |
---|---|---|---|---|---|---|---|---|
NVIDIA GeForce RTX 3080
![]() |
GA102-200-KD-A1 | 10 GB | 8704 | 272 | 96 | 1440 MHz | 1710 MHz | 1188 MHz |
NVIDIA GeForce RTX 3090
![]() |
GA102-300-A1 | 24 GB | 10496 | 328 | 112 | 1395 MHz | 1695 MHz | 1219 MHz |
NVIDIA RTX A6000
![]() |
48 GB | 10752 | 336 | 112 | 1455 MHz | 1860 MHz | 2000 MHz | |
NVIDIA RTX A40
![]() |
48 GB | 10752 | 336 | 112 | 1305 MHz | 1755 MHz | 1812 MHz | |
NVIDIA GeForce RTX 3080 Ti | GA102-250-KD-A1 | 12 GB | 10240 | 320 | 112 | 1365 MHz | 1665 MHz | 1219 MHz |
NVIDIA CMP 90HX | GA102-100-A1 | 10 GB | 8704 | 272 | 96 | 1440 MHz | 1710 MHz | 1188 MHz |