This week, Amazon (AMZN) Web Services, or AWS, kicked off re:Invent 2023 — its biggest conference of the year — with a show of force.
After a rock band banged out a rendition of “My Hero” by Foo Fighters on Tuesday morning, AWS CEO Adam Selipsky unveiled a deluge of announcements primarily related to AI. Star partners, including Nvidia (NVDA) CEO Jensen Huang and Anthropic CEO Dario Amodei, dropped by for cameos.
The surprise appearance of Nvidia’s Huang was especially notable, as he joined Selipsky on stage to announce the expansion of AWS and Nvidia’s partnership. The alliance will be characterized by a number of initiatives, including what the companies call “the first cloud AI supercomputer,” run with NVIDIA’s Grace Hopper Superchip and AWS’s UltraClusters. The cluster computing system allows customers to access thousands of powerful GPUs.
AWS will also be the first to host Nvidia’s DGX Cloud, an AI-training-as-a-service platform capable of training generative AI and large language models (LLMs) beyond one trillion parameters.
Huang seemingly nodded to Amazon and AWS’s longtime support, telling the audience that “AWS was the world’s first cloud provider to really recognize the importance of GPU-accelerated computing.”
AWS and Selipsky emphasized the company’s partnership with OpenAI competitor Anthropic, which makes the LLM Claude. In September, Amazon announced plans to invest up to $4 billion in Anthropic, whose backers also include Google (GOOG, GOOGL), Salesforce (CRM) Ventures, and Zoom (ZM) Ventures.
Selipsky took a swing at Microsoft-backed (MSFT) Open AI and its recent boardroom melodrama, referencing the events of the “past 10 days” and saying that “you don’t want a cloud provider that’s beholden primarily to one model provider.”
Amodei, who co-founded Anthropic, came on stage to discuss AI’s applications in medicine, law, and finance, including a new partnership with Pfizer (PFE).
“We’ve seen a lot of organic adoption,” said Amodei. “One example is in the biomedical space … AI can advance medical science and bring about life-saving drugs around the world.”
Additionally, AWS revealed a number of processors and chips focused on optimizing compute power and energy efficiency, including the Graviton4 and Trainium2. The latter is geared towards training large foundation models, or FMs, and LLMs faster while reducing cost.
The company’s AI announcements snowballed from there. AWS, for example, introduced a product geared towards AI safety called Guardrails. It’s now integrated into Bedrock, the company’s platform for building generative AI apps.
There was also an eye towards providing services to enterprise customers of varying AI needs. For one, the new AI chatbot Amazon Q is geared towards business customers and trained on 17 years’ worth of data from AWS.
In the end, it was all AI, all the time — and it’s not just AWS claiming territory in the AI war. Ultimately, AWS is looking to AI for its next phase of growth, which has been a subject of concern for company watchers over the last few quarters. In Q3 2023, AWS slightly missed Wall Street’s net sales expectations, coming in at $23.06 billion, against the $23.13 billion estimate.
There was good news, however. AWS sales were up 12% year over year, as Amazon CEO Andy Jassy told analysts in the company’s earnings call that AI represents an opportunity worth “tens of billions” for AWS.
But AWS isn’t planning for AI to be its quick fix. “This isn’t the work of a few months or even a year,” Selipsky emphasized, adding that Amazon and AWS are taking “long-term view of what success in generative AI really means.”
“This long-term focus is a good thing,” he added. “It’s the only predictable way to profoundly change what’s possible.”
Allie Garfinkle is a Senior Tech Reporter at Yahoo Finance. Follow her on X, formerly Twitter, at @agarfinks and on LinkedIn.
Click here for the latest technology news that will impact the stock market.
Read the latest financial and business news from Yahoo Finance
