AMD Removed RAM Encryption From Ryzen — Without Telling You
AMD silently disabled TSME on consumer Ryzen CPUs via AGESA firmware. Cold boot attacks now work. Intel still ships TME on all chips. ...
MLX + JACCL: Distributed AI Training Over Thunderbolt 5
Apple shipped JACCL at WWDC 2026 — an open-source collective communication library that enables distributed ...
Surface RTX Spark Dev Box: Run 120B Models Locally
Microsoft’s Surface RTX Spark Dev Box runs 120B models locally with 128 GB unified memory ...
Android XR Catalyst Program: Apply by June 30 for Free Dev Kit
Google's Android XR glasses launch October 14. Apply for the Catalyst Program by June 30 ...
MRC Protocol: How OpenAI’s Open Network Standard Saves GPU Clusters
OpenAI's Multipath Reliable Connection (MRC) protocol sprays packets across hundreds of paths, cutting GPU cluster ...
NVIDIA RTX Spark: Running a 120B Model Locally
NVIDIA RTX Spark puts 128 GB of unified memory in a laptop and runs 120B LLMs locally. Here is what the GB10 superchip ...
AkiraOS: WebAssembly Brings Docker-Style OTA to Microcontrollers
AkiraOS runs sandboxed WebAssembly apps on ESP32, nRF5x, and STM32 over-the-air without reflashing firmware — ...
Microsoft Project Solara: Android AI Agent OS at Build 2026
Microsoft just built a new operating system for AI agents — and it runs on ...
Quantinuum IPO at $14.3B: What Quantum Developers Actually Get
Quantinuum prices its IPO today at $14.3B. Here is what developers actually get: Guppy (free, ...
Nvidia RTX Spark: The CUDA Laptop for Local AI
Nvidia RTX Spark runs CUDA natively on a laptop. Here's what developers need to know ...