Vishal Ganeriwala

huge shoutout to @nvidia for lending engineers to help triage our security advisories 🛡️🦞 open source security hits different when GPU companies show up to help

NVIDIA has released Nemotron Nano 9B V2, a small 9B reasoning model that scores 43 on the Artificial Analysis Intelligence Index, the highest yet for <10B models Nemotron 9B V2 is the first Nemotron model pre-trained by @NVIDIA. Previous Nemotron models have been developed by post-training on Meta Llama models. Architecture & Training: The model uses a hybrid Mamba-Transformer architecture. NVIDIA pre-trained a 12B parameter base model and applied post-training with a range of techniques including RLHF and GRPO. The final 9B size was pruned from this model and re-trained with the base model as a teacher. Small-model frontier: with only 9B parameters, Nemotron Nano 9B V2 is placed ahead of Llama 4 Maverick on our leaderboard, equal to Solar Pro 2 with reasoning and trails just behind gpt-oss-20B (high). Along with this model, NVIDIA released a 6.6-trillion token subset of their pre-training data for public use on @huggingface Key model details: ➤ 128k token context window ➤ Supports reasoning and non-reasoning modes (with ‘/no_think’ settings in the system prompt) ➤ Released under the NVIDIA Open Model License, and not additionally covered by Meta’s Llama license like prior Nemotron models - this means that there is no limitation on use by large companies or requirement to keep ‘Nemotron’ in the name of derivative models ➤ No serverless inference providers are yet serving the model, but it is available now on Hugging Face for local inference or self-deployment See below for our full analysis and key announcement links from NVIDIA 👇

👀 @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. Congratulations to OpenAI on launching GPT-5, packed with enhanced reasoning and coding power. 🎉

CoreWeave​ is planning on being part of DGX Cloud Lepton marketplace to meet AI demand. Learn more & join Early Access - hubs.la/Q03rN9D60

@jeffboudier It has been great to partner with you and your team. HuggingFace totally rocks and we think we can help developers connect with cloud providers

Packed DGX Cloud sessions at #GTC Zoey and Robert doing deep dive on DGX Cloud

The open source DeepSeek-R1 model is now available as an NVIDIA NIM microservice preview on build.nvidia.com to help developers securely experiment with its advanced AI reasoning capabilities.

HiPerGator AI 2.0 has arrived ✨ UF is the first university to receive the latest @nvidia technologies — a $24 million upgrade to one of the world's most powerful supercomputers that will propel #GatorNation to new heights. news.ufl.edu/2025/01/fastes…

Is it just me or the iOS 18 features are just underwhelming? I mean arrange icons on screen was their hero feature

@djfeller Amazing

It took some time, but got some good photos from #solareclipe2024. Totality was 2x as long as 2017. And nearing solar maximum, was more corona and flare activity. Amazing to see. Plus, great location in Greers Ferry, AR Full blog: chaoticnebula.com/2024/04/17/the… #Astrophotography

@reillyusa @Dr_JCF Hope everything is ok my friend, sending prayers

Unequivocally, today has been the hardest day of our lives. Jane and I will be forever indebted to the consummate skill of @Dr_JCF and his wonderful team. From the bottom of our hearts, thank you. 🙏❤️

I spoke with an ISV working on some cutting edge AI today on why they are building on top of NVIDIA GPUs. Their answer. - When you buy and build on an NVIDIA GPU. You are not just buying a GPU, you are getting 1000s of framework and the full ecosystem around it. #GTC24

@DrJimFan @DrJimFan - incredible session today morning. I can’t wait for autonomous robots foundational models that you and your team are building #GTC24

Today is the beginning of our moonshot to solve embodied AGI in the physical world. I’m so excited to announce Project GR00T, our new initiative to create a general-purpose foundation model for humanoid robot learning. The GR00T model will enable a robot to understand multimodal instructions, such as language, video, and demonstration, and perform a variety of useful tasks. We are collaborating with many leading humanoid companies around the world, so that GR00T may transfer across embodiments and help the ecosystem thrive. GR00T is born on NVIDIA’s deep technology stack. We simulate in Isaac Lab (new app on Omniverse Isaac Sim for humanoid learning), train on OSMO (new compute orchestration system to scale up models), and deploy to Jetson Thor (new edge GPU chip designed to power GR00T). Announced in Jensen's keynote, Project GR00T is a cornerstone for the “Foundation Agent” roadmap of the newly founded GEAR Lab. At GEAR, we are building generally capable agents that learn to act skillfully in many worlds, virtual and real. See if you can spot "GEAR" in the video ;) Join us on the journey to land on the moon.

This is it …AI inferencing will never be the same . NVIDIA NIMs #GTC24

The first big news of the day is the launch of ⁦@nvidia⁩’s latest GPU architecture, Blackwell, named after mathematician David Blackwell. #GTC24

Bright and early …ready to go #GTC24

I would pay for AI tool that can help me with the expense reporting to Concur. At the end of the trip, I just want to take one picture of all the receipts and send to the tool. The tool does the magic, I verify in concur and hit done. How hard can this be?

@matteastwood @NVIDIAGTC All good and hope to see you at GTC

@VishalG @NVIDIAGTC I am connected with Mel. I don't know if she has any more room at the inn, but I will ask. Hope all is well Vishal.

I'm feeling like I should really try and get to @NVIDIAGTC next month!

@chrisfleck I was waiting for your post tell us that there is Citrix receiver for Vision Pro 🤣

VisionPro first impressions: Pros: Amazing new experience, high resolution, great large screen and multi display viewing. 3D photo viewing. No bothersome VR type claustrophobia. Cons: Not comfortable enough for extended wear time. Can’t be shared easily. Price.. It’s a V1 but a great taste of the future.