Latest news with #Trainiumtwo
Business Insider
01-08-2025
- Business
- Business Insider
Amazon.com CEO says will continue to have very deep partnership w/ Nvidia
CEO Jassy says: 'Today, so much of the cost is in training because customers are really training their models and trying to figure out how get applications into production. But in at scale, know, 80 to 90% of the cost will be an inference. Because you only train periodically, but you're spitting out predictions and and inferences all the time. And so what they're gonna care a lot about is they're gonna care about the compute and the hardware they're using. And, you know, we have a very deep partnership with NVIDIA (NVDA) and and will for as long as I can foresee, but we we we saw this movie in the CPU space with Intel where are hankering for better price performance. And so, you know, we built just like in in the CPU space where we built our own custom silicon and building Graviton, which is about 40% more price performance than the other leading x86 processors. We've done the same thing on the custom silicon side in AI with trading And our second version of Tranium two is really you know, it's it's become the backbone of Anthropix, you know, next cloud models they're trading on top of. And it's become the the backbone of of Bedrock and the inference that we do. I think a lot of the inference is about 3040% better price performance than the other GPU providers out there right now. And we're already working on our third version of trading as well. So, I think a lot of the you know, compute and the inference is gonna ultimately be run on top of Trainium two. And I think that that price performance will matter to people as they get to scale.' Comments taken from Q2 earnings conference call Q/A. Elevate Your Investing Strategy:
Business Insider
01-08-2025
- Business
- Business Insider
Amazon.com CEO says AI chip Trainium two landing capacity in larger quantities
CEO Jassy states: 'Our custom AI chip Trainium two is landing capacity in larger quantities and has improved impressively emerged as the backbone for Anthropix newest generation cloud models and many of our most essential offerings like Amazon Bedrock. We've also launched Amazon EC two instances powered by NVIDIA Grace Blackwell Superchips AWS's most powerful NVIDIA GPU accelerated instance. Second, in Bedrock, we've recently added Anthropix Cloud four and it's the fastest growing model ever in Bedrock. We've also continued to see strong adoption of Amazon Nova, our own frontier model, and it's now the second most popular foundation model in Bedrock. New features in Nova allow customers to customize their Nova models in ways they can't on other foundation models. Allowing organizations to infuse these models with their unique expertise while optimizing for cost and speed. As people have become excited about building agents, they're realizing they lack the tools to build them. In May, we released strands, an open source way to more easily build agents that's taken off with a wide range of customers already 2,500 stars in GitHub and over 300,000 downloads on PYPI. Customers are also struggling with deploying agents into production in a secure and scalable way. It's holding up enterprises scaling agents. To help solve that problem, Bedrock just released agent core. AgentCore is a set of building blocks that gives customers the industry's first secure serverless runtime to provide both synchronous and asynchronous execution. Agent identity and boundaries, a memory service, a gateway that translates services to MCP compatible interfaces, built in code execution and web browser tools, and an observability service. Customers are excited about Agent Core, and it frees them up to start deploying agents more expansively. Third, you're starting to see AWS release more powerful applications at the top layer of the AI stack.'
