AMD launches Helios, its first rack-scale AI system, securing Microsoft, Meta, OpenAI, and Oracle as customers in a bold challenge to Nvidia's 95% data center GPU market dominance, with analysts predicting AMD could capture up to 25% market share.
AMD launches Helios, a rack-scale AI system challenging Nvidia's Vera Rubin, securing Microsoft, Meta, and OpenAI as partners for deployment despite a higher price tag of $5–$5.5 million, as surging AI infrastructure demand creates room for fierce competition.
AMD unveils its Helios Rackscale AI platform, packing 72 Instinct MI455X GPUs, 2.9 exaFLOPS of FP4 compute, and 31 TB of HBM4 memory into a single reference design, with OpenAI, Meta, Oracle, and HPE among the major partners already onboard for volume deployments expected in late 2026.
China's Z.AI completes a massive 1-gigawatt data center powered entirely by domestic chips, defying US export restrictions and signaling a major leap toward AI self-sufficiency as Beijing races to replace Nvidia semiconductors with homegrown technology.
AI inference startup General Compute secures a landmark $400 million loan from Upper90, using inference-specific SambaNova SN50 chips as collateral in a first-of-its-kind deal that signals a major shift in AI infrastructure financing and a bold challenge to Nvidia's dominance.
Alibaba's chip unit T-Head open-sources its full SAIL software stack for Zhenwu AI chips, directly challenging Nvidia's CUDA dominance by promising developers a migration path in under seven days, as China accelerates its push to break free from Western AI chip ecosystems.
Apple sends legal preservation letters to 40 former employees now at OpenAI, escalating its trade secret lawsuit that accuses the AI company of orchestrating a coordinated theft of confidential hardware and product development secrets, with key former Apple executives now holding senior roles at OpenAI, which denies all allegations.
A new open-source C++17 port of ByteDance's Depth Anything 3 model runs up to 1.31x faster than PyTorch on CPU, uses half the memory, loads 6.7x faster, and requires no Python, PyTorch, or CUDA — delivering monocular depth estimation and 3D point cloud export from a single image in a …
Global tech sovereignty inches forward as a new assessment reveals most nations remain dangerously dependent on foreign technology, with China and the US commanding dominant scores of 82% and 79% while the global average barely budges from 39% to 40% by 2030, even as semiconductor independence emerges as a rare …
Nvidia-backed cloud startup Fireworks AI raises $1.5 billion at a $17.5 billion valuation after surging to over $1 billion in annualized revenue — a fivefold jump — fueled by surging corporate demand for cheaper, open-source AI models as it processes 40 trillion tokens daily at costs up to ten times …