AIWadmin
NVIDIA says Vera Rubin racks handle agents at 30x the efficiency
NVIDIA claims its Vera Rubin racks deliver up to 30x more throughput per megawatt for agents.
SpaceXAI takes NVIDIA’s Vera CPUs from data centers to orbit
SpaceXAI will run agentic AI on NVIDIA Vera CPUs on the ground and in orbit.
AMD chips fall to memory controller scrambling exploit
A new exploit scrambles DRAM addressing to break CPU isolation on AMD chips.
Anonymous Ox Alpha model has the internet guessing its maker
A free mystery model on OpenRouter has the internet guessing who built it.
Cloudflare OS gives teams a sandboxed AI work platform
Cloudflare open-sources a sandboxed AI platform it says is safe for vibe coding.
Tiny London lab’s 27B agent outshines frontier rivals on paper replication
A DeepMind alumni startup says its 27B-parameter Faraday agent beat frontier models at reproducing scientific papers.
Harvey previews its first in-house legal model for long case runs
Harvey releases Tenet, a research-preview model post-trained on long-horizon legal work from an open Kimi K3 base.
Berkeley researchers run a 753B model on a single workstation GPU
Berkeley and UT Austin researchers serve the 753B GLM-5.2 on a single workstation GPU with a new open-source…
Sixfold memory cost jump pushes OVHcloud to reprice its servers
Memory costs six times higher than a year ago push OVHcloud to raise server prices as much as…
Google opens a compiler that runs AI on encrypted data
Google's open-source HEIR compiler rewrites pre-trained models to run inference on encrypted data.
UK weighs canceling a $400M NHS deal as one region opts out
Greater Manchester refuses Palantir's NHS platform as the UK government weighs canceling a $400M contract.
OpenAI launches a research blog on AI and the future of power
OpenAI's new Strategic Futures team launches AI Futures, a blog on how transformative AI could reshape political power.
OpenAI does a U-turn on California’s AI transparency law
OpenAI, which once lobbied against California's SB 53, now wants the state to add monitoring and cybersecurity safeguards…
Safety group grades five AI labs on rogue model shutdown plans
A new scorecard from Guidelight AI Standards finds no major lab has published a complete plan for containing…
AWS opens a benchmark that tests agents on live cloud work
AWS's new open-source benchmark runs agents against real resources in disposable cloud accounts.