Hardware

AMD has released Lemonade, a high-performance, open-source server for running LLMs locally. It uniquely utilizes both GPUs and NPUs for efficient AI processing.

The upcoming 'Wine' compatibility layer is getting a revolutionary rewrite, moving Windows system call translation into the Linux kernel for massive speed gains. This architectural shift could significantly benefit not just gamers but also AI developers running Windows-native tools on Linux systems.

A viral video claiming to show an iPhone 17 Pro running a massive 400B parameter LLM has stunned the AI community. This leap in on-device power could redefine privacy and app capabilities.

George Hotz's tinygrad project has launched the Tinybox, a $15,000 personal AI supercomputer designed to run massive language models offline. Packed with six AMD GPUs, it offers a powerful, open-source alternative to Nvidia's dominance and cloud-based AI.

NVIDIA and Hugging Face have released a landmark dataset and foundational models for healthcare robotics, aiming to equip robots with physical intelligence.

A sudden shutdown of Qatar's helium production has put the global semiconductor industry on high alert. This critical gas is essential for manufacturing the advanced chips that power the AI revolution, and the supply chain now faces a critical two-week crunch.

IBM is shrinking the world of speech AI with its new Granite 4.0 1B model. This compact, multilingual powerhouse is designed for edge devices, enabling fast, private, and offline transcription without sending your data to the cloud.

Hugging Face and NXP have released a practical guide for running advanced robotics AI on small, embedded devices. Learn how they use custom data and optimization.

Apple has unveiled a redesigned iPad Air, but the real story is under the hood: the new M4 chip. This brings elite-level AI performance to a more accessible device.

NVIDIA and Hugging Face have partnered to streamline the deployment of advanced Vision Language Models (VLMs) on the Jetson platform, bringing powerful AI to the edge.

Stealth startup Taalas has just unveiled a new AI processor promising a staggering 17,000 tokens/second, an order of magnitude faster than current systems. This potential breakthrough in LLM inference could dramatically lower costs and enable truly real-time AI applications.

While Nvidia's GPUs grab headlines, the silent surge in memory demand is reshaping AI infrastructure costs. The insatiable appetite of large models for HBM is the new bottleneck.