Joseph B George (JBG)

Accuracy and precision remain extremely critical to what we do in science, health, economics and other research. It has to be – it's what your constituents / peers need when it comes to scientific work. Why? In scientific computing, precision is directly connected to trust. Whether researchers are modeling climate systems, designing aerospace platforms, accelerating drug discovery, or developing AI-driven scientific workflows, the quality of the answer matters more than simply getting an answer quickly. Even tiny numerical inaccuracies can compound across billions of calculations into materially different outcomes. In many scientific environments, numerical stability is not optional — it is foundational to the validity of the result. Read more here: x.com/jbgeorge/statu… #HPC #AI #HighPerformanceComputing #ArtificialIntelligence #AIforScience #Science #Research #MachineLearning

The AI industry is racing toward lower precision everything — FP8, FP6, FP4, and beyond. And for many AI workloads, that makes perfect sense. Lower precision can dramatically improve throughput, reduce power consumption, and lower infrastructure cost. But across national laboratories, scientific research institutions, government agencies, and enterprise HPC environments, one reality remains unchanged: Achieving a correct result is the end goal. Period. FP64 remains extremely critical to what we do in science, health, economics and more. It has to be – it's what our constituents and peers need when it comes to scientific work. Learn more in my new article: "Newsflash: Getting a Wrong Answer FAST... Is Still Getting a Wrong Answer" here: x.com/jbgeorge/statu… #HPC #HighPerformanceComputing #ArtificialIntelligence #AIforScience #FP64 #MathematicalPrecision #ScientificComputing #MachineLearning #DataCenter #AIInfrastructure #Supercomputing #Technology #Innovation #DataScience

Tremendous evening at the @The Austin Forum on Technology & Society last night, where I was on a panel with @Dell’s Onur Celebioglu and @Google's Bill Magro discussing Supercomputing, Cloud Computing, and AI Factories - and the Importance of Speed and Scale. Great questions from Google's Jay Boisseau which drove strong conversation, as well as great questions from the audience. A few items I want to reiterate that I highlighted yesterday: 1. Supercomputing has been driving innovation for decades - and you use a lot of it everyday. I specifically called out the Weather app on your phone - it's able to predict what happens 3 days from now for a reason. 2. When it comes to AI, the best thing you / anyone / everyone can do is _get_started_now_. Find an LLM of your choice, and start working on any idea you have. That's it - don't worry about code, technical items, etc just yet - we need "big idea" people that will get out there and work on their idea. You can't break these AI tools, so don't worry about that - don't wait for a class, just start... and learn with the rest of us. 3. Bill Magro mentioned NotebookLM as a great place to start - I'm a HUGE fan of NotebookLM - I wrote an article about AI solving network optimization in the telecommunications industry recently, and NotebookLM created a fully AI-generated podcast for me. Super awesome - highly recommend. 4. Lastly, my view is that our industry / our world does best when supercomputing + AI + cloud WORK TOGETHER vs working separately. And with quantum computing around the corner, I expect we will add quantum to the equation. It's a GREAT time to be working in all of these space - so jump in! Thanks, Jay and the Austin Forum team - great work getting us together to talk about a super important topic! #Supercomputing #CloudComputing #AIFactories #AI #Quantum #Austin #AMD #Google #Dell

Austin Tech Friends! This Tuesday, join me, alongside @Google CTO William Magro, and @DellTech CTO Fellow Onur Celebioglu, as we delve into "Supercomputing, Cloud Computing, and AI Factories: The Importance of Speed and Scale" It's another great topic hosted by @AustinForum on Technology & Society and by Google's Advanced Computing Strategist Jay Boisseau. On the agenda - Differences, commonalities, convergence of the tech and approaches - Current applications - and the future applications we see on the horizon - Challenges that we expect to face as an industry - And so much more WHEN: Tuesday, December 2, 2025, 6:15 PM to 7:45 PM WHERE: Google Austin, 500 West 2nd Street, Austin, TX Click the link below to reserve your free spot and to learn more - and I hope to see you there! austinforum.org/events/decembe… #AMD #Google #Dell #AI #Cloud #Supercomputing #CloudComputing #AIFactory #Austin #Tech

TELCO + AI: If we think through what we need to build here for both network optimization and predictive maintenance, we need a solution that can manage the telecommunications network real-time, we need a solution that can anticipate maintenance / potential failure, and we need a solution that is heavily automated. Enter AI. linkedin.com/pulse/ai-telco… #artificalintelligence #AI #telecommunications #telco #networkoptimization #predictivemaintenance

@elonmusk Meet the ‘woolly mouse’: why scientists doubt it’s a big step towards recreating mammoths nature.com/articles/d4158…

Please make a miniature pet wooly mammoth

AI IN TELECOMMUNICATIONS - "If we think through what is needed for a telecommunications solution that can allow for predictive maintenance and network optimization: - We need a solution that can manage the telecommunications network real-time - We need a solution that can anticipate maintenance / potential failure - We need a solution that is heavily automated. Enter AI." #Telecommunications #Telco #ArtificialIntelligence #AI linkedin.com/pulse/ai-telco…

DDN featured in Times Square! Huge day! @DDN #NYC #TimesSquare #ArtificialIntelligence #NASDAQ

Join us at DDN on Feb 20! REGISTER HERE: lnkd.in/gYvQBTTC #ArtificialIntelligence #AI @ddnstorage

On February 20, 2025, you can have a front-row seat to one of the biggest product breakthroughs in AI data intelligence.... #ArtficialIntelligence #AI #NVIDIA #datacenter #cloud

A Blackstone investment values AI data company DDN at $5 billion trib.al/H3ASQ3o