AI & Compute roundup: Astera Labs adds fabric memory for AI racks
Leo X plugs DRAM into AI accelerator fabrics for KV cache offload; AWS also open-sourced an email-style AI agent manager.
Tag Ollama
RTX Spark N1X Brings Up to 128GB of Unified Memory to October Laptops
NVIDIA set October for RTX Spark, split the N1X into a 32GB and a 128GB tier, and shipped a router that spreads agent work across the PCs already on a home network.