We’ve partnered with @AMD, @Broadcom, @Intel, @Microsoft, and @NVIDIA, to release Multipath Reliable Connection (MRC), a new open networking protocol that helps large AI training clusters run faster and more reliably, with less wasted GPU time.
https://t.co/AiV952AJXs
The @AMDServer CPU with integrated HBM3 still amazes me that it was released, and wasn't heavily utilized or marketed. But the @Azure team putting it in HPC instances makes ABSOLUTE sense.
Really nice work by @Signal_65 articulating the #HPC performance and customer value prop of @Azure HBv5 VMs. ~5x speedup v. alternatives for CFD is huge!!
For HPC workloads, time-to-solution is the metric that matters. Faster iterations mean more design cycles, earlier forecasts, and shorter research timelines.
In our testing, the new @Azure HBv5 instances completed an OpenFOAM CFD run with a 100M-cell mesh in 261 seconds. The prior HBv4 generation took 1,278 seconds. That is a 4.9x speedup on a single workload.
The driver is integrated HBM3 keeping all 368 cores on the @AMDServer CPU fed with data, not waiting on DDR5 channels.
Full report: https://t.co/K7AeaScVc1
@nicholasmalaya@nicholasmalaya this a useful and rational take, so TY. But given that standard (a) nothing more “sovereign” v. any USA CSP site b) chips still made in Taiwan+Korea (not bad, just no diff) and (c) workforce dev @ 1 DOE site moving needle v 100’s of existing CSP/neo-cloud sites?
@nicholasmalaya@glennklockwood@nicholasmalaya Because people figuring out the DCs that can be leased today won’t be able to accommodate the GPU supercomputers coming down the pike.
No one becomes a clinician to do paperwork, but it's becoming a bigger and bigger administrative burden, taking time and attention away from actually treating and supporting patients.
That’s why we’re introducing Microsoft Dragon Copilot, the industry’s first AI assistant for clinical workflow.
Today, we are excited to announce Thinking Machines Lab (https://t.co/Pe0uB8MxVN), an artificial intelligence research and product company. We are scientists, engineers, and builders behind some of the most widely used AI products and libraries, including ChatGPT, https://t.co/CTXVZKNH1c, PyTorch, and Mistral. Our mission is to make artificial intelligence work for you by building a future where everyone has access to the knowledge and tools to make AI serve their unique needs.
We are committed to open science through publications and code releases, while focusing on human-AI collaboration that serves diverse domains. Our approach embraces co-design of research and products to enable learning from real-world deployment and rapid iteration. This work requires three core foundations: state-of-the-art model intelligence, high-quality infrastructure, and advanced multimodal capabilities. We are committed to building models at the frontier of capabilities to deliver on this promise.
If you’re interested in joining our team, consider applying here: https://t.co/fkT82CRytQ
@HPC_Guru@AMD@Microsoft@Azure@AMDServer You’re most welcome @HPC_Guru 😃 We’re thrilled to finally unveil publicly HBv5 and the custom CPU and platform we and @AMDServer built for it. We in @Azure care deeply about HPC customers using these tools to do their lives’ work. Can’t wait to get HBv5 into their eager hands!
This is fantastic! You can now run Python *locally* in Excel via the Anaconda Toolbox plugin. It complements Excel's built-in Python support, and gives you a choice for you can't use Azure. And it includes an LLM-powered assistant to help with your Python code, viz needs + more!
My quick take - Serbia may have played the best 3Q’s of contiguous bball in olympic history, and #USABasketball almost certainly played the best 1Q ever. Unbelievably high level basketball.