Now in the in the good old days when I would say hopper I would hold up a chip that's just adorable. This is Vera Rubin. This is the Vera Rubin system.100% liquid cooled. All of the cables gone. What used to take 2 days to install now takes 2 hours.Incredible. And so the manufacturing cycle time going to dramatically reduce. This is also a supercomput that is cooled by it's cooled by hot water 45 degrees which takes the pressure off of the data center takes all of that cost and all of that energy that's use ...

When the U.S. Department of Energy (DOE) switches on its next major supercomputer at the National Energy Research Scientific Computing Center (NERSC), it won’t just be adding another cluster to a lab in California. It will be putting a new model for America’s AI infrastructure into motion. Sponsored by @DellTech https://t.co/rZjuEAd84Q ...

BREAKING: xAI has committed to securing 1.2 gigawatts of power to run its supercomputer as the main energy supply.According to SpaceX President Gwynne Shotwell:“This applies to every future data center too. We're scaling up the world's largest Megapack installation already in place. It will deliver backup power capable of supporting the entire city of Memphis, plus more than enough to fully run the town of Southaven, Mississippi, where the current data center is located.” ...

Market Expansion & Partnerships - The company aims to expand its Three SC system, a new class of supercomputer, through partnerships, including those in the UAE and potentially Saudi Arabia [1] - The company has salespeople on the ground in Saudi Arabia, viewing it as an exciting opportunity, although the effort is not as advanced as in the UAE [2] - The company's success with G 42, Meta, IBM, Mayo Clinic, GlaxoSmithKline, and AI startups like Cognition facilitates entry into new markets [6] Technology & Security - The company works closely with G 42, its strategic partner, which has developed a restricted technology environment to protect IP and prevent technology transfer to China [3][4] - The company is confident that its technology is protected both physically and from use by Chinese agents, having visited facilities and seen the restricted technology environment in action [4] Government & Policy - The company believes that allowing allies and partners in the Middle East access to advanced AI compute is a good and thoughtful policy [2] - The company does significant business with the Department of Defense (DOD) and the Department of Energy (DOE), with the DOE being a leading organization for U S AI policy [8] Sales & Growth - The company acknowledges that sales in any territory take time, requiring the building of credibility and legitimacy [5] - The company has extraordinary access and is proud of its accomplishments with the government, though it always desires more support [7][8]

Supercomputing Advancements - Oak Ridge National Lab (ORNL) emphasizes its long history of major discoveries facilitated by supercomputing, leading to interesting science advancements [1] - The OLCF's Frontier, the nation's first exascale supercomputer, solves problems 100 to 1,000 times faster than before, enabling the pursuit of multiple problems simultaneously [2][3] - Discovery represents the next evolution, converging high-performance computing, quantum computing, and AI, integrating quantum into high-performance computing architectures [4] - Discovery will set a new standard for sovereign AI and leadership computing, improving upon Frontier in almost every way [5] Technology and Partnerships - Discovery will feature HPE's new GX platform and AMD's newest GPU, built in close collaboration with HPE and AMD [6][7][11] - Lessons learned from Frontier will directly translate to Discovery, making it a better machine [7] Applications and Impact - Exascale computing enables flight-scale simulations, addressing turbulent flow fields in GE engines, impacting approximately 1 million people flying on GE engines [8][9] - AI streamlines processes in the nuclear power space, enhancing the design, construction, and operation of nuclear power plants [9][10] - The combination of nuclear energy and AI has tremendous power to create good for the world [10] - Discovery increases productivity and contributes to U S competitiveness through leadership in AI and computing [11] Future Vision - The ultimate goal is to create digital twins of humans for personalized medicine, testing treatments and life choices in real-time [7][8] - Discovery is a scientific instrument that will enable faster, farther, and deeper insights than any previous machine [12]

Product Overview - DGX Spark aims to democratize AI by providing personal supercomputing capabilities to researchers, artists, and innovators [1] - DJX Spark enables developers to run state-of-the-art open-source AI models without requiring large data center GPUs [2] - DGX Spark is positioned as a supercomputer suitable for individual use, such as for every AI graduate student [3] - The product is small and portable, allowing users to run large language models (LLMs) that would otherwise be too demanding for personal laptops [5] - Spark facilitates the simulation, running, creation, and fine-tuning of vision models for edge devices [6] Target Applications - DGX Spark can serve as the brain of installations, enabling robots to generate new paintings daily [5] - Zipline utilizes simulations to test various edge cases, and DGX Spark provides the necessary computing capacity for such applications [4] - The device is suitable for running vision models in constrained compute environments [6] Accessibility and Innovation - The product aims to make robotics, computing, and open source technologies more accessible [3] - DGX Spark empowers individuals to turn imagination into discovery, making AI accessible to everyone [6]

1.21 Gigawatts of training @xAI !!! https://t.co/KF5ANCUcEUMuskonomy (@muskonomy):🚨NEWS: Today NVIDIA CEO Jensen Huang calls Elon Musk “the most world-class builder we have in the country.”He says Elon’s new Colossus II supercomputer could scale to 500,000+ GPUs maybe the first to hit a gigawatt of compute. https://t.co/X0JsUi39OT ...

Technology Upgrade - Europe's top forecaster is looking to upgrade its supercomputer [1] Weather Challenges - The upgrade responds to the challenges posed by extreme weather [1]

Isambard, Britain’s latest supercomputer, is not big enough to train the largest language models. It will, however, enable other research breakthroughs https://t.co/DT9HDZrmvj ...

AI Infrastructure & Innovation - Nvidia's AI factory, EOS, ranks as the ninth fastest supercomputer globally [1] - EOS is a large-scale Nvidia DGX super pod designed for leading-edge AI innovation [2] - EOS utilizes a full-stack architecture with Nvidia accelerated infrastructure, networking, and AI software [2] - Nvidia DGX H100 systems within EOS feature eight Nvidia H100 Tensor Core GPUs each [2] - The system is designed for training generative AI projects at high speeds [2] Enterprise AI Strategy - Enterprises can leverage AI factories like EOS to tackle demanding AI projects [3] - AI factories enable enterprises to achieve their AI aspirations [3]