Ukraine-based app developer MacPaw has partnered with Liquid AI to integrate locally hosted AI models into its products, including the Eney assistant and SetApp store.
Apple's defunct self-driving car project indirectly led to the development of its powerful Neural Engine, crucial for on-device AI processing. This technology debuted with iPhone X and now powers M-series chips, with future upgrades planned for the M7 chip in 2027.
Samsung Electronics announced Tuesday it has developed the industry's fastest Universal Flash Storage (UFS) 5.0. The new memory solution is designed to enable on-device AI, allowing AI models to run locally without a network connection, with mass production slated for Q4.
COMPUTEX TAIPEI 2026, held in Taipei from June 2-5, concluded as the largest event ever with over 110,000 attendees. The main focus was NVIDIA's new 'RTX Spark' SoC, a 13-year return to Windows PCs, integrating Arm CPUs and Blackwell GPUs for on-device Agent AI, aiming to compete with high-end Macs.
MiniCPM5-1B, a 1 billion parameter AI model, is designed for on-device deployment, supporting native tool calling and the Model Context Protocol (MCP). It outperforms comparable open-source models in its size class, offering a 128K token context window and efficient inference capabilities.
AMD unveiled the Ryzen AI Halo mini-PC on May 20, supporting up to 200 billion parameters for on-device AI processing. The Ryzen AI Max+ 395 configuration will be available for pre-order in June at Micro Center for $3,999.
Chrome is quietly installing a 4GB AI on your devices. And now, its latest version has removed the disclosure that promised to keep your data off Google’s servers.
Chrome is quietly installing a 4GB AI on your devices. And now, its latest version has removed the disclosure that promised to keep your data off Google’s servers.
