hapabapa
Google’s (NASDAQ:GOOG) (NASDAQ:GOOGL) new artificial intelligence-focused processors will be used by AI startup Anthropic as the Mountain View, California-based tech giant seeks to boost its cloud computing unit amid increased competition.
The new fifth-generation tensor processing unit chip, known as the the TPU v5e, were first unveiled in August at Google’s annual Cloud Next event. Anthropic said the TPUs will be used to power the startup’s large language model known as Claude.
Anthropic Chief Executive Officer Dario Amodei said the AI startup is “really excited about the efficiency of the TPUs” and that it is already starting to see gains, according to comments obtained by Bloomberg.
The news comes as Alphabet (GOOG) (GOOGL) continues its competition in the cloud computing space against industry heavyweights Amazon (NASDAQ:AMZN) Web Services and Microsoft’s (MSFT) Azure.
In its most recent quarter, Alphabet (GOOG) (GOOGL) said Cloud revenue rose 22.5% year-over-year to $8.41B, below Wall Street estimates, as it continued to deal with customers optimizing their spending.
Amazon (AMZN) also offers its own processors designed for cloud applications, including the AWS Graviton CPU and its AI-focused Inferentia2 chip.
Anthropic, which was founded in 2021, announced a partnership in February 2023 with Google (GOOG) (GOOGL) Cloud to deliver “responsible and reliable” AI. Google is also Anthropic’s preferred cloud provider and will continue to use its infrastructure and services.
In August, Google (GOOG) (GOOGL) said the TPU v5e is built to train LLMs. It’s also focused on bringing cost efficiency, delivering up to twice the higher training performance per dollar and up to two-and-a-half times the performance per dollar for LLMs and generative AI models compared to the previous model.
Last month, it was reported that Google (GOOG) (GOOGL) invested $2B into Anthropic.
In September, Amazon (AMZN) said it would invest up to $4B in San Francisco-based Anthropic, and take a minority stake in the company as the two parties collaborate to advance generative AI.
As part of that deal, Amazon Web Services became Anthropic’s main cloud provider for mission critical workloads, including safety research and future foundation model development. The AI-focused startup added that it intends to run the “majority of its workloads” on AWS.
