Tags: hardware acceleration

Gimlet Labs, founded by former Pixie co‑founders including Stanford adjunct professor Zain Asgar, announced an $80 million Series A led by Menlo Ventures. The startup’s “multi‑silicon inference cloud” software lets AI workloads run simultaneously across CPUs, GPUs, and high‑memory systems, promising 3‑to‑10× faster inference at the same cost and power. Partnerships with major chip makers such as NVIDIA, AMD, Intel, ARM, Cerebras and d‑Matrix support the platform, which targets large model labs and data‑center operators. The round brings total funding to $92 million and backs a team of 30 employees. Lire la suite

Qualcomm announced a new generation of Snapdragon processors designed to enable AI agents on smartphones and laptops. The chips feature upgraded Oryon CPUs, higher token‑processing speeds, and increased AI TOPS performance, allowing on‑device large language models to run faster and more efficiently. Executives highlighted the potential for personalized assistants that can access local documents, images, and apps, improving productivity and accessibility. While the hardware advances remove a major barrier, Qualcomm emphasizes that manufacturers must still integrate the AI‑agent interfaces into their products. Lire la suite