.@xai, partnering with NVIDIA and @Supermicro_SMCI, built Colossus, the world's largest liquid-cooled GPU cluster, featuring over 100,000 NVIDIA H100 GPUs and NVIDIA Spectrum-X Ethernet #networking.
Read the @VentureBeat article to learn more.
NVIDIA AI Infrastructure
11.9K posts
NVIDIA AI Infrastructure
@NVIDIAAIInfra
AI factories for the era of AI reasoning.
- 👀 @OpenAI's GPT-5 was trained on NVIDIA H100 and H200s GPUs and served on systems like NVIDIA GB200 NVL72 featuring 72 #NVIDIABlackwell GPUs and 36 Grace CPUs, connected using advanced NVIDIA NVLink and NVLink Switch computing fabrics designed for state-of-the-art AI at scale.GPT-5 is here. Rolling out to everyone starting today. openai.com/gpt-5/
00:00 - 🌏 NVIDIA and @SamsungKorea announce a new state-of-the-art AI factory to transform global intelligent manufacturing. The #AIfactory with 50,000+ NVIDIA GPUs will accelerate agentic and physical AI applications for advanced chip manufacturing, mobile devices, and robotics.
- Exciting to see Colossus, the world’s largest GPU #supercomputer, come online in record time. Colossus is powered by @nvidia's #acceleratedcomputing platform, delivering breakthrough performance with exceptional gains in #energyefficiency. Congratulations to the entire team!This weekend, the @xAI team brought our Colossus 100k H100 training cluster online. From start to finish, it was done in 122 days. Colossus is the most powerful AI training system in the world. Moreover, it will double in size to 200k (50k H200s) in a few months. Excellent
- Built with NVIDIA Grace Blackwell Ultra Superchips, #NVIDIADGX GB300 is purpose-built AI infrastructure designed for training and inferencing state-of-the-art models. Features include: ✅ Scalable up to tens of thousands of Grace Blackwell Ultra Superchips with DGX SuperPOD
- Announced during the #GTC20 Keynote, the new NVIDIA DGX SuperPOD Solution for Enterprise is a ready-made #AI supercomputer that can be up and running in just a few weeks. Learn how it's powering business transformation across industries. nvda.ws/2GQqoLG
- 📣 NVIDIA and SK Group build one of Korea’s largest AI factories, with 50,000+ NVIDIA GPUs powering the nation’s digital transformation. This groundbreaking #AIfactory will accelerate chip design, robotics, digital twins, and AI agent innovation for SK Group and partners across
- 🎉Congratulations to @nebiusai on their partnership with @Microsoft. In collaboration with NVIDIA, Nebius will deliver AI infrastructure at scale featuring #NVIDIABlackwell—unlocking massive AI capacity and fueling innovation across the industry. Learn more now. ⬇️Yesterday, we announced a multi-year agreement with @Microsoft: nebius.com/newsroom/nebiu… At the heart of it? Next-gen GPU innovation from @nvidia — including systems built on the NVIDIA Blackwell architecture. We’ve reimagined everything, from thermal design to deployment
- Our CEO Jensen Huang is hosting leaders from across the quantum industry on-stage at #GTC25 to discuss the future of #quantumcomputing at NVIDIA's first Quantum Day. 🌟 Learn more about Quantum Day and register for GTC now. nvda.ws/3Wkhbzl
- 📣 @OracleCloud has launched its first wave of liquid-cooled NVIDIA GB200 NVL72 racks, now live and available for customer use. Thousands of #NVIDIABlackwell GPUs and high-speed NVIDIA networking are being deployed across OCI’s global data centers, powering #NVIDIADGX Cloud and
- Fairwater is pioneering the future of AI. As one of the world’s most advanced AI Factories, it runs hundreds of thousands of NVIDIA GB200 NVL72 racks with NVIDIA Spectrum-X as a single, massive cluster—engineered to tackle the most demanding AI workloads efficiently, at unmatchedIf intelligence is the log of compute… it starts with a lot of compute! And that’s why we’re scaling our GPU fleet faster than anyone else. Just last year, we added over 2 gigawatts of new capacity – roughly the output of 2 nuclear power plants. And today we’re going further,
00:00 - 📣 @CoreWeave announces the deployment of #NVIDIABlackwell Ultra for AI inference at scale. The NVIDIA GB300 NVL72 systems with NVIDIA Networking delivered from @DellTech have more than one exaflop of dense AI performance and up to 40TB of fast memory in a single rack. Learn
- NVIDIA GB300 NVL72 features a power supply unit (PSU) with integrated energy storage and advanced ramp-up and ramp-down controls. ⚡ These innovations actively smooth out power fluctuations and reduce peak grid demand by up to 30%. This isn't just a cool feature; it's a
