Artificial intelligence news.
Straight from the original announcements.

DeepSeek R1

Beelink launches modular ME Pro NAS on Kickstarter
Chips and infrastructureNovember 6, 2026, 02:57

Beelink launches modular ME Pro NAS on Kickstarter

Beelink has launched the ME Pro on Kickstarter, a modular NAS available in two and four-bay configurations with an interchangeable motherboard design. The company says…

CoreWeave Doubles GB300 NVL72 DeepSeek R1 Server Throughput in MLPerf 6.0
Chips and infrastructureOctober 22, 2026, 18:27

CoreWeave Doubles GB300 NVL72 DeepSeek R1 Server Throughput in MLPerf 6.0

CoreWeave says it doubled DeepSeek R1 server-mode throughput on NVIDIA GB300 NVL72 systems compared with MLPerf 5.1, within six months. According to the company, it also…

CoreWeave Reports 6.5x DeepSeek R1 Inference Gain on NVIDIA GB300 NVL72
Chips and infrastructureOctober 22, 2026, 18:02

CoreWeave Reports 6.5x DeepSeek R1 Inference Gain on NVIDIA GB300 NVL72

CoreWeave says its new NVIDIA GB300 NVL72 instances, built on Blackwell Ultra GPUs, delivered a 6.5x performance improvement on inference for the DeepSeek R1 reasoning…

MLPerf Named Yardstick in EU AI Gigafactories Tender as Endpoints v1.0 Opens
Chips and infrastructureOctober 13, 2026, 16:25

MLPerf Named Yardstick in EU AI Gigafactories Tender as Endpoints v1.0 Opens

MLPerf has been chosen as the performance standard for bids in the European Union's call for AI Gigafactories, according to MLCommons. The organization also said MLPerf…

CoreWeave and NVIDIA Say AI Factories Must Be Proven Before Production
Chips and infrastructureOctober 2, 2026, 16:12

CoreWeave and NVIDIA Say AI Factories Must Be Proven Before Production

CoreWeave argues that having GPUs is no longer enough to be ready for AI workloads, and that the full stack must be co-designed and validated before customers depend on…

Goodfire Releases First Sparse Autoencoders for DeepSeek's 671B R1 Model
ResearchOctober 1, 2026

Goodfire Releases First Sparse Autoencoders for DeepSeek's 671B R1 Model

Goodfire has trained what it says are the first public sparse autoencoders (SAEs) on a true reasoning model, DeepSeek's 671B parameter R1, and has open-sourced them. The…