Podcast: “Supercomputing used to be about going fast in one direction. In AI, that just drives you off a cliff faster.” — @glennklockwood.com Lockwood on why flexibility, not just performance, defines next-gen infrastructure. Now at VAST. shared-everything.simplecast.com/episodes/fro...

Neat. RIKEN’s CGRA arch (research project/not product) systolic AI acceleration with FP64 HPC precision on a reconfigurable mesh. cogarchworkshop.org/assets/paper... #HPC #AI

Datacenters inside wind turbines? Aligns renewable power with flexible #AI workloads like model training but thermal challenges, latency headaches, and intermittent wind raise big questions about AI #datacenter readiness. #HPC www.vastdata.com/sharedeveryt...

McKinsey says Gen AI is stuck: lots of hype, little bottom-line impact. But banks, research firms, and others are quietly breaking through with AI agents and something called the "agentic mesh." Is this AI’s big pivot moment? www.vastdata.com/sharedeveryt...

Countries now see data as territory to protect. If the U.S. doesn't tune in, it might get tuned out. www.vastdata.com/sharedeveryt...

Just my hydrangeas.

In drug development, where timelines are long, costs are high, and many bets fail, any tool that helps reduce risk is worth keeping. But only if we’re honest about what it can’t do yet. #AI - www.vastdata.com/sharedeveryt...

All things, with one exception, have to end After years of sharing #HPC news and insights, it's time to log off It was fun while it lasted and I remain grateful to all of you who followed and interacted Stay curious, stay kind! — @hpcguru.bsky.social signing off #HPC #Farewell

Antoine Bordes, ex-head of AI at Meta FAIR now leading AI at Helsing, discusses intelligent defense, data-driven warfare, and ethical AI integration in military applications. www.vastdata.com/sharedeveryt...

TACC's Dan Stanzione on Horizon and how supercomputing is evolving beyond raw compute: real-time AI workflows, power efficiency, quantum futures, and why data rules the stack. www.vastdata.com/sharedeveryt...

An in-depth chat with NERSC chief architect Nick Wright exploring how the forthcoming NVIDIA Vera Rubin CPU+GPU Doudna machine redefines supercomputing to meet real-time demands of science. www.vastdata.com/sharedeveryt... #HPC #AI

Take a moment to dive into #DigitalTwin infrastructure and future work blending #HPC and #AI. I'm joined this week for the Shared Everything podcast by Wes Brewer from Oak Ridge National Lab and Adrian Jackson from EPCC. youtu.be/s-Y4SQ4etA4?...

Some insight into how CoreWeave is replacing Hollywood’s render farms with AI-native data engines that turn every frame into a dataset. www.vastdata.com/sharedeveryt...

Here’s what it’ll take for Nvidia and other US chipmakers to flog AI chips in China

Everyone’s talking about NVIDIA Dynamo... disaggregation, smarter routing, KV cache tricks: all solid moves. But behind the GPU inference magic, there’s a deeper infrastructure story nobody’s mentioning. www.vastdata.com/sharedeveryt...

GenAI’s infrastructure crisis of costly inference and rigid stacks is driving an open, Linux-like shift. Modular, community-built solutions are already rewriting datacenter rules as we saw with Red Hat's announce today (among other cues) www.vastdata.com/sharedeveryt...

My HS senior made the rounds to her old elementary school to see her 4th grade handprint on the wall and her youngest sister. I know it’s cliche ya’ll but time..it really does fly.

Remember hearing about Koomey's Law? With AI in the mix, this datacenter energy figure just keeps more interesting. Here's my chat with Koomey filmed this week www.vastdata.com/sharedeveryt...

Lip-Bu Tan: Intel’s new – and maybe last – CEO Analysis and commentary here on Intel's future and its new boss www.nextplatform.com/2025/03/13/l...

Is object storage taking over AI training? What comes after parallel file systems? How do we speed long-context #AI? Moderating a convo with AI Infra Architect at Microsoft, Glenn Lockwood and Jeff Denworth at #GTC. Have breakfast with me! -https://events.vastdata.com/nvidia-gtc-breakfast-microsoft

If you are at #GTC this session has to be a priority: 1. Because of #HPC celebs involved - @glennklockwood.com @nicolehemsoth.bsky.social & Jeff Denworth 2. The important topic for #AI & #supercomputing - the future of #IO, #data and filesystems (or not) events.vastdata.com/nvidia-gtc-b...

A Q&A that details why I am moving from reporting on the future to helping build it. Big change but for great reasons. www.vastdata.com/blog/nicole-...

I really wish these folks would stop naming their projects after perfectly good science fiction shows.

A terminal case of #AI fever: It's not just one hyperbolic billionaire – the entire industry is chasing the #AI dragon Million #GPU clusters, #gigawatts of power – the scale of AI defies logic www.theregister.com/2024/12/19/s... #HPC via @tobiasmann.bsky.social @theregister.com

Just how deep is Nvidia's CUDA moat really?

Many refreshing things about being here but top of the list this week are all the book recommendations found via people I don’t follow directly. Open to more, it’s winter and I have a birthday spa weekend ahead. Just finished Fayne, took sci-fi break but ready for more. #Booksky (I’m guessing)

China gorging on silicon before Uncle Sam slams the door

It’s that time of year again…

Imagine working for a company that puts out ads like this. "We're keeping you around to harvest skills then dumping you."

Meta’s new datacenter even looks like a chip. From www.nextplatform.com/2024/12/05/t...

Just when we think we've seen it all, along comes a PR agency weaponizing Creative Commons licenses Insist the content of announcements and quotes from execs are CC licensed, requiring attribution -- fine -- but crucially, a link through to the source Thus ensuring clients get links in articles

the thing about using ChatGPT as a search engine is that we literally solved the problem of reliable search engines a decade ago, proceeded to destroy that in search of ad revenue, then reinvented it again, but somehow weirdly reliant on reddit shitposts

AWS says AI could disrupt everything – and hopes it will do just that to Windows

More than half of the server CPU capacity installed at AWS is now homegrown, and it won’t be long before this is true for AI accelerators. $AMZN $NVDA $AMD $INTC www.nextplatform.com/2024/12/03/a... #AWSreinvent

Amazon promises 4x faster AI silicon in 2025, turns Trainium2 loose on the net

The cost of Gelsinger’s ambition proves too much for Intel

If China can’t get HBM, then that memory will be available for other uses, which will benefit Nvidia, AMD, and those making their own AI accelerators. More supply will be available for them to play with, which is the real upside (whether Biden admin realized it or not)…

“Intel still has two-thirds of the share of the X86 server market, and the X86 server recession that has plagued Intel and bothered AMD – and that we think was caused in part by a shift in focus on spending for AI systems-is largely over… “ www.nextplatform.com/2024/12/02/w...

Interesting visualisation by Visual Capitalist on the results of TOP500’s latest #supercomputer ranking from a country perspective. More info: www.visualcapitalist.com/ranked-top-c...

Amazon founder Jeff Bezos joined Samsung in a $700 million bet on Tenstorrent, valuing the #AI chip startup with ambitions of taking on Nvidia at about $2.6 billion www.bloomberg.com/news/article... #HPC #RISCV

Pat Gelsinger is out as Intel CEO, and the debate begins anew about breaking Intel up. Consider TSMC and the US government when calculating your odds. www.nextplatform.com/2024/12/02/w...

Intel CEO Pat Gelsinger retiring effective today. www.intc.com/news-events/...

Nothing like coming back to the real world after break to the immediate #awsreinvent news deluge…

Wood Brothers at the Orange Peel Asheville this Friday and Avett Brothers coming to MerleFest in April. I can get through this winter 🙌

If you are from the American Midwest this is the only acceptable receptacle for green bean casserole

Tried a starter pack on #semiconductors - please let me know who else should be in! go.bsky.app/AwKDHi9

Fascinating.

I was trying to explain US #Thanksgiving traditions and food to a British colleague and it was hard so I found an explainer

HPE is upgrading its supercomputer lineup across compute, storage, and interconnects in 2025. We go through the details. $HPE $AMD $NVDA $INTC www.nextplatform.com/2024/11/26/h...

I recently saw an amazing Navajo rug at the National Gallery of Art. It looks abstract at first, but it is a detailed representation of the Intel Pentium processor. Called "Replica of a Chip", it was created in 1994 by Marilou Schultz, a Navajo/Diné weaver and math teacher. 1/n