Jump to content

Annapurna Labs

From Wikipedia, the free encyclopedia
(Redirected from Trainium)

Annapurna Labs is an Israeli microelectronics company. Since January 2015 it has been a wholly owned subsidiary of Amazon.com. Amazon reportedly acquired the company for its Amazon Web Services division for US$350–370M.[1][2]

History

[edit]

Annapurna Labs, named after the Annapurna Massif in the Himalayas, was co-founded in 2011[3] by Bilic "Billy" Hrvoje, a Bosnian Jewish refugee, Nafea Bshara, an Arab Israeli citizen,[4][5] and Ronen Boneh with investments from the independent investors Avigdor Willenz, Manuel Alba, Andy Bechtolsheim, the venture capital firm Walden International, Arm Holdings,[6] and TSMC. Board members include Avigdor Willenz, Manuel Alba, and Lip-Bu Tan, the CEO of Walden International and executive chairman of Cadence Design Systems.[7]

The first product launched under the AWS umbrella was the AWS Nitro hardware and supporting hypervisor in November 2017.[8] Following on from Nitro, Annapurna developed general-purpose CPUs under the Graviton family and machine-learning ASICs under the Trainium and Inferentia brands.[9][10][11]

In November 2024 Annapurna announced their second generation Trainium 2 intended for training AI models. Based on their internal testing, Amazon are claiming "a 4-times performance increase between Trainium 1 and Trainium 2".[12]

See also

[edit]
  • AWS Graviton - an ARM based CPU developed by Annapurna Labs for exclusive use by Amazon Web Services.

References

[edit]
  1. ^ "Amazon to buy Israeli start-up Annapurna Labs". Reuters. 22 January 2015. Archived from the original on 2015-09-15. Retrieved 2015-01-24.
  2. ^ "Amazon buys secretive chip maker Annapurna Labs for $350 million". ExtremeTech. Archived from the original on 2020-12-14. Retrieved 2015-01-24.
  3. ^ Clark, Greg; Bensinger, Dan (2016-01-06). "Amazon Enters Semiconductor Business With Its Own Branded Chips". The Wall Street Journal. Archived from the original on 2023-10-21. Retrieved 2024-05-21.
  4. ^ "Annapurna Labs: AWS' Secret Sauce". Retrieved 2019-12-09.
  5. ^ Rebecca Kopans. "If you can dream it, you can do it" (PDF). Archived from the original (PDF) on 2019-12-09. Retrieved 2019-12-09.
  6. ^ Kristen Lisa. "AWS and ARM: Working together to re-invent the cloud". Archived from the original on 2019-12-09. Retrieved 2019-12-09.
  7. ^ "Semiconductors fueling Cloud!". semiwiki.com. Archived from the original on 2016-03-03. Retrieved 2015-01-24.
  8. ^ Liguori, A (2018). "The Nitro Project–Next Generation AWS Infrastructure" (PDF). Hot Chips: A Symposium on High Performance Chips. Institute of Electrical and Electronics Engineers (IEEE). Archived (PDF) from the original on 18 May 2022. Retrieved 13 October 2023.
  9. ^ Tarasov, Katie (12 August 2023). "How Amazon is racing to catch Microsoft and Google in generative A.I. with custom AWS chips". CNBC. Archived from the original on 13 October 2023. Retrieved 13 October 2023.
  10. ^ Bass, Dina (2023-02-21). "Amazon's Cloud Unit Partners With Startup Hugging Face as AI Deals Heat Up". Bloomberg News. Archived from the original on 2023-05-22. Retrieved 2024-05-21.
  11. ^ Nellis, Stephen (2023-02-21). "Amazon Web Services pairs with Hugging Face to target AI developers". Reuters. Archived from the original on 2023-05-30. Retrieved 2024-05-21.
  12. ^ Michael Acton; Tim Bradshaw. "Amazon ready to use its own AI chips, reduce its dependence on Nvidia". Ars Technica. Retrieved 12 November 2024.
[edit]