Google accelerates custom AI chip development, shortening TPU release cycles to two or more per year
Amin Vahdat, Google's VP and Chief Architect of AI Infrastructure, stated that Google has shortened the release cycle for its TPUs from one model every two years to one model every year, and even two models per year last year. He added that this pace will further accelerate in the future to maintain a lead in the AI competition. Google plans to invest up to $205 billion this year in building AI infrastructure.
Source:Nikkei Asia · Source Link
Disclaimer: This content reflects only the author’s personal views and does not constitute any investment or financial advice. If you discover any content that violates regulations,Click to Report
24H Trending
-
1
SanDisk: How Does the Storage Giant Outperform Bitcoin and Gold Amidst the AI Wave?
-
2
HBOT Token: The Governance Core and Future Outlook of the Hummingbot Ecosystem
-
3
A Look Back at 2023: Which Cryptocurrencies Were Predicted to Surge, and How Did They Actually Perform?
-
4
What is TWT? An In-depth Analysis of Trust Wallet Token's Value and Prospects
-
5
VV Token Trading Guide: Understanding Virtual Versions (VV) and How to Buy It
-
6
HTX (Huobi) Exchange Plagued by Controversy: An In-depth Analysis of its Strengths, Weaknesses, and Challenges
-
7
GMC Coin Value Analysis: Distinguishing Multiple Projects with the Same Name and Long-Term Investment Considerations
-
8
What is BIL? An Analysis of Billions Network (BILL) and its Market Status
-
9
DYDX (dYdX) Analysis: The Evolution and Current State of a Decentralized Perpetual Contract Platform
-
10
ADX (AdEx Network) Project Analysis and Future Prospects
Markets Today
Recommended Reading












