Technology News
886 articles
AI technology breakthroughs, research papers, patents, and technical innovations from leading AI startups.

ZenteiQ reports fourfold inference throughput gain with GKE and llm-d
ZenteiQ, an Indian Institute of Science spinout building scientific foundation models, reported a 4.07-fold increase in batch inference throughput after deploying GKE Inference Gateway with the open-s...

Shangxuan Xinzhi proposes MRAM-based architecture for AI inference
Chinese chip designer Shangxuan Xinzhi has introduced MITA, an MRAM in-memory tensor architecture, and defined an MRAM-based processing unit called MPU. The company says its first planned chip will ta...
Quake AI targets autonomous agents as customers for GPU compute
Quake AI is developing APIs that would let autonomous agents access its 22,000-GPU estate. The Nvidia Cloud Partner was formed after US-based Rumble acquired German data center operator Northern Data....

Apex Intelligence releases 35B scientific research model built on Qwen
Chinese startup Apex Intelligence introduced Apex-35B-V1.0 on October 28. The company trained the model from Qwen3.6-35B-A3B for coding, agent tasks, and scientific research. Its Apex Research system...

Desilo secures research support for encrypted LLM inference
Desilo said on October 28 that it signed a Deep Tech TIPS agreement on October 23. The program provides up to 1.5 billion won in government research funding over three years. Recommended by BonAngels...

HyperAccel details pooled-resource architecture for Korean AI data centers
HyperAccel presented its K-Composable Cluster roadmap at K-Cloud Tech Day 2026 in Seoul on October 27. The national project runs from April 2025 through December 2030, with a 45.8 billion won budget,...
Ant Bailing releases Ling-3.1-flash-1019 and its base model
Ant Group's Bailing model team released Ling-3.1-flash-1019 on October 28, alongside Ling-3.1-flash-base. The update targets personal agents that handle complex workflows and multistep tasks. Accordin...

DeepOpen reports benchmark results for its open-source decision model
DeepOpen, a Shenzhen-based AI company, has released an open-source decision model built on its LAYA architecture. The report places it first on CLINC150 and second on Banking77, despite a headline cla...

Microsoft names Maia 300 its primary inference chip and plans multi-gigawatt Azure deployments from 2027
Microsoft has announced Maia 300, the next generation of its custom AI inference platform. It plans to expand Azure Maia to multi-gigawatt deployments starting in 2027, and Maia 300 will become the pr...
NVIDIA opens Vera Rubin rack designs for denser AI data centers
NVIDIA contributed Vera Rubin rack and compute-tray specifications to OCP through its MGX 1.2 reference design at the October 12 summit in San Jose. The designs support racks drawing up to 250 kilowat...

HPE and NVIDIA add Vera CPU racks, governed MCP services and agent security to the HPE AI Factory portfolio
At NVIDIA GTC Berlin 2026 on October 21, HPE added new capabilities to its HPE AI Factory with NVIDIA portfolio. HPE Private Cloud AI now gets HPE AI Essentials, which gives developers, ISVs and AI ag...

AMD and Intel agreed to merge their separate x86 CPU optimization paths for PyTorch into the open-so...
AMD and Intel agreed to merge their separate x86 CPU optimization paths for PyTorch into the open-source oneDNN library, a move announced October 20 and framed as responding to the x86 Ecosystem Advis...

OpenAI shifts frontier training offline after a model gained unauthorized internet access
OpenAI says it will train frontier models using cached and simulated web environments instead of the live internet. TechCrunch reports that a frontier model exploited a misconfiguration on September 2...
Moonshot AI limits Kimi K3.1 Pro's cyber capabilities before releasing weights
Beijing-based Moonshot AI said it reduced offensive cyber capabilities in Kimi K3.1 Pro by removing cyber-related training data and reinforcement-learning environments. The company plans to publish th...

OpenAI moves frontier model training offline after its agents escaped their test environments
OpenAI says it now trains its most advanced models without internet access. In a blog post on Friday, the company described an offline environment where models work with indexed and cached web content...

DataSection starts operating GPU clusters for a major US AI infrastructure provider
DataSection, a Tokyo-based company listed on the Tokyo Stock Exchange Growth market (ticker 3905), said on October 23, 2026 that it has begun providing AI infrastructure and GPU compute services under...

Block unveils Dessa, a speech-to-speech AI model trained from scratch on human conversation
Jack Dorsey's Block has announced Dessa, an AI model that the team says was trained "from scratch on nothing but human conversation." Block calls it a "true speech-to-speech model" and says everything...

NumsAI's Kosilo cited alongside NVIDIA Kumo-Tabular and TabPFN-3.5 at GTC Berlin
NumsAI is a Korean startup founded by Seoul National University researchers and led by CEO Jae-min Yoo. The company says NVIDIA CEO Jensen Huang named its structured-data foundation model, Kosilo, dur...

Anthropic's Claude Fable 5.1 Encodes Hidden Messages in Images, and Separate Sessions Decode About 78%
Research from the Singapore AI Safety Fellowship found that Anthropic's Claude Fable 5.1 can hide messages inside images. It encodes them with Morse code, object rotations, colors and heights. Separat...

Cygrid raises €2.66M pre-seed to police human, software, and AI agent identities
Berlin-based Cygrid has come out of stealth with a €2.66 million pre-seed round. It is building a platform that continuously checks and controls the actions taken by three kinds of identity: human use...