#inference
articles
Companies
IBM Strikes $240M Deal with $8.3B Startup Together AI for Massive Inference Infrastructure
The multi-year agreement will see IBM’s cloud host Nvidia-powered systems dedicated to running open-source AI models fro...
Technology
Memory, not compute: Majestic Labs bets LPDDR6 pooling can crack Nvidia's AI wall
Tel Aviv startup Majestic Labs says AI inference is bottlenecked by memory, not GPUs — and its Arm-based Prometheus serv...
Companies
Analysts Turn Bullish on CoreWeave, DigitalOcean and Nebius as AI Inference Demand Accelerates
Truist and Baird upgrade or initiate coverage on three neocloud providers, citing a shift from AI training to inference ...
Technology
Infinity Lands $15M to Chip Away at Nvidia’s CUDA Moat with Universal AI Inference Software
The startup, backed by Touring Capital and researchers from OpenAI and Anthropic, is building a self-optimizing inferenc...