2-bit quantization is the only reason a 320-billion-parameter model appears in the same sentence as a laptop. Z.ai shipped GLM-5.3-Flash in August 2026 under an MIT licence, and within days the community had squeezed the full-precision checkpoint from 641.64 GB down to 108.72 GB. The headlines that followed said the same thing in different words: […]
RTX Spark superchip laptops go on sale this autumn, and they are not another spin on the Copilot+ badge. Nvidia has taken the Grace CPU and Blackwell GPU pairing it built for desk-side AI boxes, shrunk it into a chassis as thin as 14mm, and persuaded six of the largest PC makers in the world […]
Apple announced the M5 Ultra on 25 August 2026 alongside a new Mac Studio: an up-to-80-core GPU with a Neural Accelerator in every core, up to 512GB of unified memory at 1.2TB/s, a 32-core Neural Engine and a media engine rated for 33 simultaneous streams of 8K ProRes 422. This breakdown separates the peak-theoretical claims from the measured workload figures, works out what 512GB actually buys you when you run large models locally, checks the 8K video claim against a real timeline, lists the prices and ship dates, and sets out the questions Apple did not answer.
Local LLM mobile app integration enables businesses to add private, responsive and offline artificial intelligence capabilities to native iOS and Android applications.
A practical guide to AirLLM, the open-source inference library that reduces memory requirements for very large open models by loading layers one at a time.
