Intel and Nvidia Recast AI Platforms
Partnership targets heterogeneous client and data center systems
OpenAI’s $180B chip bet hits a wall
Broadcom wants Microsoft to commit before money moves
Tenstorrent goes general-purpose in an age of AI specialization
Company says one system can do the whole inference job.
IBM’s latest AI processor is GPU-free
It does have 32 NPUs, however.
AI Vector Vision reframes computer vision
Deterministic, vector-based representation.
Taalas bets on hard‑wired models to beat GPUs at inference
Toronto start-up Taalas fixes trained LLMs directly in silicon.
Vaire Computing bets on reversible chips
Sunnyvale, CA, start-up tackles AI power problem at the physics level.
SK Hynix backs Semidynamics AI chips
Strategic investment accelerates memory-first AI inference platform.
Socionext’s custom XPU and AI data center chiplet platform
Partners up with Arm's CPU.
Rebellions’ Rebel100 NPU
Designed for servers and pods.
Fujitsu joins the AIP crowd with an NPU
Targeted at inference servers using advanced Japanese semi production.
AI processor architectures are evolving almost as fast as LLMs are
AGI will not be realized with just more clever software.