AI agent infrastructure is maturing but facing reliability challenges, as FlashAttention-4 achieves 71% utilization on NVIDIA's Blackwell B200 GPUs while agentic AI systems grapple with architecture q…
The FEX Emulator, a Valve-backed project for running x86/x86_64 software on ARM64 systems, released its FEX 2607 update with optimizations for yet-to-be-released 256-bit SVE2 ARM processors, improveme…
Anthropic released Claude Science, a multi-agent AI workbench for reproducible genomics, proteomics, and cheminformatics pipelines. The beta app runs on existing Claude models and integrates over 60 c…
NVIDIA Research introduced HORIZON, a hands-free agent framework for hardware design that treats RTL development as repository-level code evolution, achieving 100% completion across every evaluated be…
NVIDIA announced Halos for Robotics on June 22, 2026, a full-stack safety system for robotics and physical AI. Agility is the first company using Halos for humanoid robots in factories and warehouses …
OpenAI's Stargate UK data center project, announced with a £30 billion investment figure, has been paused after it emerged that OpenAI never visited a key planned site and that roughly £20 billion of …
NVIDIA and academic researchers introduced ASPIRE, a self-improving robotics framework that achieves 31% zero-shot success on LIBERO-Pro long tasks. The system uses a coordinator-actor architecture wi…
AMD's Instinct MI355X achieves roughly 80% of NVIDIA B200 throughput on GLM-5.2 inference at over 2x lower cost, delivering more than 2x better tokens-per-dollar after hand-tuning the ROCm software st…
Raja Koduri's startup OXMIQ raised $35 million in Series A funding on July 1 to license its OxCore GPU architecture, aiming to let developers run CUDA code on non-NVIDIA hardware without changes. The …
Over $610 billion in AI hardware capital expenditure was committed globally in a single week, including South Korea's $550B memory fab investment, Japan's $6B for AI model development, and Qualcomm's …
Two former Harvard students, dissatisfied with expert claims that a dedicated inference chip is physically impossible, founded Etched and raised $800 million to build an ASIC for transformer model inf…
OpenAI previewed GPT-5.6 Sol on June 26, 2026, but has not yet released the model. Early unverified reports indicate the model initially underperforms Claude Opus in short sessions but surpasses it af…
Austria's MUSICA supercomputer, featuring 1,088 NVIDIA H100 GPUs delivering 45.11 petaflops, was inaugurated on July 3, 2026, marking an eightfold increase over previous systems. The EUR 45 million sy…
Firefly Aerospace will launch an NVIDIA Jetson AI computing platform into lunar orbit on its Blue Ghost Mission 2, enabling autonomous AI processing on the Moon without relying on Earth commands. The …
Wafer served GLM5.2 on AMD MI355X GPUs at 2626 tokens per second per node with over 2x lower cost than NVIDIA Blackwell, achieving 213 tok/s single stream. The company used MXFP4 quantization via AMD …
Palantir CEO Alex Karp used a CNBC interview to promote the company's new 'AI sovereignty' manifesto and an expanded partnership with NVIDIA offering open-weight AI models to U.S. government agencies,…
The Conference on Robot Learning (CoRL) 2026 will be held in Austin, Texas, from November 9 to 12, chaired by Yuke Zhu and Peter Stone. The event will focus on robot foundation models, real-robot rein…
The Linux Foundation launched Project Akrites on June 25 to defend critical open source software against AI-enabled cyber threats, establishing a shared Security Incident Response Team backed by 19 fo…
Running state-of-the-art large language models locally requires either a $50,000+ multi-GPU rig or a software-driven pipeline decomposition approach, as memory bandwidth—not compute—is the primary bot…
NVIDIA released the Nemotron-3-Ultra-550B-A55B-NVFP4 model, a 550-billion-parameter large language model with 55 billion active parameters using NVFP4 quantization, under the OpenMDW-1.1 license. The …