NVIDIA GA107

GA107

GN20-P0-A1

SM Diagram

NVIDIA's GA107 GPU uses the Ampere architecture and is made using a 8 nm production process at Samsung. With a die size of 200 mm² and a transistor count of 8,700 million it is a medium-sized chip. GA107 supports DirectX 12 Ultimate (Feature Level 12_2). For GPU compute applications, OpenCL version 3.0 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 2560 shading units, 80 texture mapping units and 32 ROPs. Also included are 80 tensor cores which help improve the speed of machine learning applications. The GPU also contains 32 raytracing acceleration cores.

Further reading: Ampere Architecture Whitepaper

Graphics Processor

Released: Apr 12th, 2021

GPU Name: GA107

Codename: NV177

Architecture: Ampere

Foundry: Samsung

Process Size: 8 nm

Transistors: 8,700 million

Density: 43.5M / mm²

Die Size: 200 mm²

Package: FCBGA-1358

Graphics Features

DirectX: 12 Ultimate (12_2)

OpenGL: 4.6

OpenCL: 3.0

Vulkan: 1.3

CUDA: 8.6

Shader Model: 6.7

WDDM: 3.1

Tensor Cores: 3rd Gen

RT Cores: 2nd Gen

NVENC: 7th Gen

NVDEC: 5th Gen

PureVideo HD: VP11

VDPAU: Feature Set K

Render Config

Shading Units: 2560

TMUs: 80

ROPs: 32

SM Count: 20

FP16 Units: 2560

FP64 Units: 64

INT32 Units: 1280

Tensor Cores: 80

RT Cores: 32

SFUs: 320

TPCs: 10

GPCs: 2

Tex L1 Cache: 64 KB per SM

L1 Cache: 128 KB per SM

L2 Cache: 2048 KB

Max. TDP: 250 W

All Ampere GPUs

NVIDIA GPU Architecture History

2024 Blackwell
2023 Hopper
2022-2024 Ada Lovelace
2020-2024 Ampere
2018-2022 Turing
2017-2020 Volta
2016-2021 Pascal
2014-2019 Maxwell 2.0
2014-2017 Maxwell
2013-2015 Kepler 2.0
2012-2018 Kepler
2010-2016 Fermi 2.0
2010-2013 VLIW Vec4
2010-2016 Fermi
2007-2013 Tesla 2.0
2006-2010 Tesla
2003-2013 Curie
2003-2005 Rankine
2001-2003 Kelvin
1999-2005 Celsius
1998-2000 Fahrenheit

Graphics cards using the NVIDIA GA107 GPU

Name	Chip	Memory	Shaders	TMUs	ROPs	Base Clock	Boost Clock	Memory Clock
NVIDIA GeForce RTX 3050 4 GB	GA107-140-A1	4 GB	2048	64	32	1545 MHz	1740 MHz	1500 MHz
NVIDIA GeForce RTX 3050 Ti Mobile		4 GB	2560	80	32	735 MHz	1035 MHz	1500 MHz
NVIDIA GeForce RTX 3050 Mobile	GN20-P0-A1	4 GB	2048	64	32	1065 MHz	1343 MHz	1500 MHz
NVIDIA RTX A4 Mobile		4 GB	2048	64	32	1297 MHz	1770 MHz	1750 MHz
NVIDIA A16 PCIe		16 GB	1280	40	32	1312 MHz	1755 MHz	1563 MHz
NVIDIA A2		16 GB	1280	40	32	1440 MHz	1770 MHz	1563 MHz
NVIDIA GeForce RTX 2050 Mobile		4 GB	2048	64	32	735 MHz	1245 MHz	1750 MHz
NVIDIA RTX A2000 Embedded	GA107-980-A1	4 GB	2560	80	32	607 MHz	1177 MHz	1500 MHz
NVIDIA GeForce RTX 3050 8 GB GA107	GA107-150-A1	8 GB	2560	80	32	1552 MHz	1777 MHz	1750 MHz
NVIDIA RTX A1000 Mobile		4 GB	2048	64	32	630 MHz	1140 MHz	1375 MHz
NVIDIA GeForce RTX 3050 Max-Q Refresh 6 GB		6 GB	2560	80	32	622 MHz	990 MHz	1375 MHz
NVIDIA GeForce RTX 3050 Mobile Refresh 6 GB		6 GB	2560	80	32	1237 MHz	1492 MHz	1750 MHz
NVIDIA GeForce RTX 3050 Max-Q		4 GB	2048	64	32	712 MHz	1057 MHz	1375 MHz
NVIDIA RTX A500		4 GB	2048	64	32	1440 MHz	1770 MHz	1750 MHz
NVIDIA GeForce RTX 2050 Max-Q		4 GB	2048	64	32	832 MHz	1155 MHz	1475 MHz
NVIDIA GeForce RTX 3050 Max-Q Refresh 4 GB		4 GB	2048	64	32	757 MHz	1125 MHz	1375 MHz
NVIDIA A2 PCIe		16 GB	1280	40	32	1440 MHz	1770 MHz	1563 MHz
NVIDIA RTX A1000 Mobile 6 GB		6 GB	2560	80	32	652 MHz	1140 MHz	1375 MHz
NVIDIA GeForce RTX 3050 Ti Max-Q		4 GB	2560	80	32	735 MHz	1035 MHz	1375 MHz
NVIDIA GeForce RTX 3050 Mobile Refresh 4 GB		4 GB	2560	80	32	652 MHz	1207 MHz	1750 MHz
NVIDIA GeForce RTX 3050 6 GB	GA107-325-K2-A1	6 GB	2304	72	32	1042 MHz	1470 MHz	1750 MHz