Image created with Flux Pro v1.1 Ultra. Image prompt: NVIDIA, GPU heat-sink fins suggested by densely aligned rows of small green bananas, subtle PCB hints, photorealistic, editorial, minimal, high detail, 3:2 landscape
NVIDIA continues to lead on open-sourcing pretraining data — Nemotron-CC-v2 has dropped! https://x.com/ZeyuanAllenZhu/status/1962119316427706828
Jensen on NVIDIA Q2 Earnings Call: “”Our new robotics computing platform, Thor, is now available. Thor delivers an order of magnitude greater AI performance and energy efficiency than NVIDIA’s AGX Orin. It runs the latest generative and reasoning AI models at the edge in real https://x.com/TheHumanoidHub/status/1961342309209100670
Nvidia launched Jetson AGX Thor, a $3,499 chip for real-time physical AI It uses a 2,560-core Blackwell GPU, 96 fifth-generation Tensor cores, and 128GB of memory to deliver up to 2,070 FP4 teraflops of AI compute https://x.com/adcock_brett/status/1962184408246415687
NVIDIA’s Jetson AGX Thor, a $3,499 ‘robot brain,’ is now available. Powered by a Blackwell GPU with 128GB memory, it delivers up to 2,070 FP4 teraflops in 130W. Early adopters include Boston Dynamics, Agility, and Figure—pushing humanoid robotics into a new era. 🤖✨ https://x.com/StarSnap_1/status/1960153258389053561
NEW: Google is talking to several GPU cloud providers about putting its tensor processing units in their data centers.
The push to expand in the data centers of Nvidia-focused cloud providers is a new strategy for Google. https://x.com/anissagardizy8/status/1963228123144819167
ZeroGPU on 🤗 HF Spaces enables anyone to build delightful ML demos, benefitting from powerful compute. But, due to its serverless nature, it is hard to optimize these demos. That CHANGES today 🪖 Use AoT compilation to melt our ZeroGPU servers 🔥 Details ⬇️ https://x.com/RisingSayak/status/1962844485118996545
ZeroGPU on Hugging Face enables anyone to build and deploy AI apps dynamically allocates and releases NVIDIA H200 GPUs as needed But, due to its serverless nature, it is hard to optimize these apps Now use AoT compilation to melt ZeroGPU servers on Hugging Face for vibe https://x.com/_akhaliq/status/1962920105186115621
NVIDIA Blackwell GPUs are incredible, but requires knowing the hardware to get the most out of it. This blog post series aims to demystify what it takes to get peak performance out of this sophisticated device!”” / X https://x.com/clattner_llvm/status/1961491323875455029
So, Nvidia is doing ablation on 13B model for 10T token only to show their 4-bit (NVFP4) training is stable? https://x.com/eliebakouch/status/1962805948184998064
Rest the world finally picking up on Google selling TPUs externally.
In the Accelerator model, we’ve discussed details about it over a month ago months regarding both TPUv7 Ghostfish and TPUv8 Sunfish / Zebrafish https://x.com/dylan522p/status/1963355683170246659
🍁 In collaboration with @NVIDIAAIDev, @RedHat_AI, and @VectorInst, vLLM is hosting a meetup in Toronto September 25th! Come hear about project update, distributed inference, EAGLE spec decode, and FlashInfer! https://x.com/vllm_project/status/1963736578674893071
GPT-OSS uses MXFP4 quantization (which MLX now supports). There are two FP4 formats circulating right now: MXFP4 and NVFP4 (NV for Nvidia). From looking at how GPT-OSS uses MXFP4, it is somewhat suboptimal. I’m thinking NVFP4 will be the more commonly used format in the https://x.com/awnihannun/status/1961500133990043967
