Ampere (microarchitecture) explained

Ampere
Designfirm:Nvidia
Manuf1:TSMC
Manuf2:Samsung
Process:TSMC N7
Codename:GA10x
Products-Desktop1:GeForce RTX 30 series
Products-Hedt1:RTX A series
Products-Server1:A100
Directx-Version:DirectX 12 Ultimate (Feature Level 12_2)
Direct3d-Version:Direct3D 12.0
Shadermodel-Version:Shader Model 6.8
Opencl-Version:OpenCL 3.0
Opengl-Version:OpenGL 4.6
Cuda-Version:Compute Capability 8.6
Vulkan-Api:Vulkan 1.3
L1-Cache:192KB per SM
128KB per SM
L2-Cache:2MB to 6MB
Pcie-Support:PCIe 4.0
Encoders:NVENC
Support Status:Supported

Ampere is the codename for a graphics processing unit (GPU) microarchitecture developed by Nvidia as the successor to both the Volta and Turing architectures. It was officially announced on May 14, 2020 and is named after French mathematician and physicist André-Marie Ampère.[1] [2]

Nvidia announced the Ampere architecture GeForce 30 series consumer GPUs at a GeForce Special Event on September 1, 2020.[3] [4] Nvidia announced the A100 80 GB GPU at SC20 on November 16, 2020.[5] Mobile RTX graphics cards and the RTX 3060 based on the Ampere architecture were revealed on January 12, 2021.[6]

Nvidia announced Ampere's successor, Hopper, at GTC 2022, and "Ampere Next Next" (Blackwell) for a 2024 release at GPU Technology Conference 2021.

Details

Architectural improvements of the Ampere architecture include the following:

Chips

Comparison of Compute Capability: GP100 vs GV100 vs GA100[11]

GPU featuresNvidia Tesla P100Nvidia Tesla V100Nvidia A100
GPU codenameGP100GV100GA100
GPU architecturePascalVoltaAmpere
Compute capability6.07.08.0
Threads / warp323232
Max warps / SM646464
Max threads / SM204820482048
Max thread blocks / SM323232
Max 32-bit registers / SM655366553665536
Max registers / block655366553665536
Max registers / thread255255255
Max thread block size102410241024
FP32 cores / SM646464
Ratio of SM registers to FP32 cores102410241024
Shared Memory Size / SM64 KBConfigurable up to 96 KBConfigurable up to 164 KB

Comparison of Precision Support Matrix[12] [13]

Legend:

Comparison of Decode Performance

Concurrent streamsH.264 decode (1080p30)H.265 (HEVC) decode (1080p30)VP9 decode (1080p30)
V100162222
A10075157108

Ampere dies

DieGA100[14] GA102[15] GA103[16] GA104[17] GA106[18] GA107[19] GA10B[20] GA10F
Die size826mm2 628mm2 496mm2 392mm2 276mm2 200mm2??
Transistors54.2B28.3B22B17.4B 12B 8.7B??
Transistor density65.6 MTr/mm245.1 MTr/mm244.4 MTr/mm244.4 MTr/mm243.5 MTr/mm243.5 MTr/mm2??
Graphics processing clusters87663221
Streaming multiprocessors12884604830201612
CUDA cores1228810752768061443480256020481536
Texture mapping units512336240192120806448
Render output units192112969648323216
Tensor cores512336240192120806448
RT coresN/A8460483020812
L1 cache24MB10.5MB7.5MB6MB3MB2.5MB3MB1.5MB
192KB
per SM		128KB per SM192KB
per SM		128KB
per SM
L2 cache40MB6MB4MB4MB3MB2MB4MB?

A100 accelerator and DGX A100

The Ampere-based A100 accelerator was announced and released on May 14, 2020. The A100 features 19.5 teraflops of FP32 performance, 6912 FP32/INT32 CUDA cores, 3456 FP64 CUDA cores, 40 GB of graphics memory, and 1.6 TB/s of graphics memory bandwidth.[21] The A100 accelerator was initially available only in the 3rd generation of DGX server, including 8 A100s.[22] Also included in the DGX A100 is 15 TB of PCIe gen 4 NVMe storage, two 64-core AMD Rome 7742 CPUs, 1 TB of RAM, and Mellanox-powered HDR InfiniBand interconnect. The initial price for the DGX A100 was $199,000.

Products using Ampere

Products using Ampere (per Chip)
Type GA10B GA107 GA106 GA104 GA103 GA102 GA100
GeForce MX seriesGeForce MX570 (mobile)
GeForce 20 seriesGeForce RTX 2050 (mobile)
GeForce 30 seriesGeForce RTX 3050 Laptop
GeForce RTX 3050
GeForce RTX 3050 Ti Laptop		GeForce RTX 3050
GeForce RTX 3060 Laptop
GeForce RTX 3060		GeForce RTX 3060
GeForce RTX 3060 Ti
GeForce RTX 3070 Laptop
GeForce RTX 3070
GeForce RTX 3070 Ti Laptop
GeForce RTX 3070 Ti
GeForce RTX 3080 Laptop		GeForce RTX 3060 Ti
GeForce RTX 3080 Ti Laptop		GeForce RTX 3070 Ti
GeForce RTX 3080
GeForce RTX 3080 Ti
GeForce RTX 3090
GeForce RTX 3090 Ti
Nvidia Workstation GPUsRTX A1000 (mobile) RTX A2000 (mobile)
RTX A2000 		RTX A3000 (mobile)
RTX A4000 (mobile)
RTX A4000
RTX A5000 (mobile) 		RTX A5500 (mobile) RTX A4500
RTX A5000
RTX A5500
RTX A6000
Nvidia Data Center GPUsNvidia A2
Nvidia A16 		Nvidia A10
Nvidia A40 		Nvidia A30
Nvidia A100
Tegra SoCsAGX Orin
Orin NX
Orin Nano

See also

External links

Notes and References

  1. Web site: NVIDIA's New Ampere Data Center GPU in Full Production. NVIDIA. Newsroom. NVIDIA Newsroom Newsroom.
  2. Web site: NVIDIA Ampere Architecture In-Depth. May 14, 2020. NVIDIA Developer Blog.
  3. Web site: NVIDIA Delivers Greatest-Ever Generational Leap with GeForce RTX 30 Series GPUs . Nvidia Newsroom . en-US . September 1, 2020 . April 9, 2023.
  4. Web site: NVIDIA GeForce Ultimate Countdown . Nvidia . en-US.
  5. Web site: NVIDIA Doubles Down: Announces A100 80GB GPU, Supercharging World's Most Powerful GPU for AI Supercomputing . Nvidia Newsroom . en-US . November 16, 2020 . April 9, 2023.
  6. Web site: NVIDIA GeForce Beyond at CES 2023. NVIDIA.
  7. Web site: I.7. Compute Capability 8.x . Nvidia . en-US . September 23, 2020.
  8. Web site: Bosnjak . Dominik . September 1, 2020 . Samsung's old 8nm tech at the heart of NVIDIA's monstrous Ampere cards . SamMobile . en-US. September 19, 2020.
  9. Web site: Delgado . Gerardo . September 1, 2020 . GeForce RTX 30 Series GPUs: Ushering In A New Era of Video Content With AV1 Decode . Nvidia . en-US . April 9, 2023.
  10. Web site: Morgan . Timothy Prickett . May 29, 2020 . Diving Deep Into The Nvidia Ampere GPU Architecture . The Next Platform . en-US . March 24, 2022.
  11. Web site: NVIDIA A100 Tensor Core GPU Architecture: Unprecedented Accerlation at Every Scale . Nvidia . en-US . September 18, 2020.
  12. Web site: NVIDIA Tensor Cores: Versatility for HPC & AI. NVIDIA.
  13. Web site: Abstract. docs.nvidia.com.
  14. Web site: NVIDIA A100 Tensor Core GPU Architecture . NVIDIA Corporation . en-US . April 29, 2024.
  15. Web site: NVIDIA GA102 GPU Specs . TechPowerUp . en-US . April 29, 2024.
  16. Web site: NVIDIA GA103 GPU Specs . TechPowerUp . en-US . April 29, 2024.
  17. Web site: NVIDIA GA104 GPU Specs . TechPowerUp . en-US . April 29, 2024.
  18. Web site: NVIDIA GA106 GPU Specs . TechPowerUp . en-US . April 29, 2024.
  19. Web site: NVIDIA GA107 GPU Specs . TechPowerUp . en-US . April 29, 2024.
  20. Web site: NVIDIA AGX Orin Series Technical Brief v1.2 . NVIDIA Corporation . April 29, 2024.
  21. News: Nvidia's first Ampere GPU is designed for data centers and AI, not your PC. Tom Warren. James Vincent. May 14, 2020. The Verge.
  22. News: Smith. Ryan. May 14, 2020. NVIDIA Ampere Unleashed: NVIDIA Announces New GPU Architecture, A100 GPU, and Accelerator. AnandTech.
  23. Web site: Igor . Wallossek . February 13, 2022. The two faces of the GeForce RTX 3050 8GB . Igor's Lab . February 23, 2022 . igor-ga107.
  24. Web site: Shilov. Anton. September 25, 2021. Gainward and Galax List GeForce RTX 3060 Cards With GA104 GPU. Tom's Hardware. September 23, 2022.
  25. News: Tyson. Mark. February 23, 2022. Zotac Debuts First RTX 3060 Ti Desktop Cards With GA103 GPU. Tom's Hardware. September 23, 2022.
  26. Web site: WhyCry . October 26, 2022 . ZOTAC launches GeForce RTX 3070 Ti with GA102-150 GPU . VideoCardz . en-US . May 21, 2023.

This article is licensed under the GNU Free Documentation License. It uses material from the Wikipedia article "Ampere (microarchitecture)".

Except where otherwise indicated, Everything.Explained.Today is © Copyright 2009-2024, A B Cryer, All Rights Reserved. Cookie policy.