UALink
UALink (Ultra Accelerator Link) is an open industry standard for scale-up interconnect between AI accelerators that lets up to 1,024 accelerators inside a single pod read and write each other's memory directly…
Open Source AI
Explore Open Source AI through related topics and the articles other pages reference most.
Explore articles
Browse subtopics: AI Compute
Articles that also belong to these categories. Counts cover all of Open Source AI.
All articles (512)Large Language Models (183)AI Models (132)Developer Tools (127)Chinese AI (92)AI Agents (68)AI Companies (48)AI Infrastructure (46)Mixture of Experts (43)Multimodal AI (31)NVIDIA (31)Robotics (31)AI Code Generation (28)Machine Learning (26)Natural Language Processing (23)AI Inference (22)Meta AI (20)Small Language Models (19)AI Tools & Products (18)Generative AI (18)Data & Datasets (16)Diffusion Models (15)Video Generation (15)Speech & Audio AI (14)Image Generation (13)Information Retrieval (13)Reasoning Models (13)Computer Vision (12)AI Research (11)Google (11)Research Organizations (11)Training & Optimization (11)AI Safety (10)Embodied AI (10)Microsoft (10)Reinforcement Learning (10)Artificial Intelligence (9)Google DeepMind (8)MLOps (8)Robot Hardware (8)AI for Science (7)Voice AI (7)OpenAI (6)AI Benchmarks (5)AI Hardware (5)Deep Learning (5)Drug Discovery (5)Enterprise AI (5)Healthcare AI (5)Model Evaluation (5)AI Policy & Regulation (3)Algorithms (3)Anthropic (3)Conversational AI (3)Humanoid Robots (3)Interpretability (3)People (3)Software Development (3)World Models (3)AI Compute (2)AI Ethics (2)AI in Gaming (2)Agentic Commerce (2)Mathematics (2)Model Architecture (2)Neural Networks (2)Physical AI (2)Programming Languages (2)Transformer Models (2)AI Events (1)AI Incidents & Controversies (1)Autonomous Vehicles (1)Consumer AI Devices (1)Data Science (1)Finance AI (1)Music & Audio Generation (1)Quantum Computing (1)Robotics Companies (1)
Showing 1-2 of 2 articles
Ultra Ethernet
Ultra Ethernet is an open networking specification that reworks Ethernet into a high performance fabric for large AI training clusters and high performance computing, giving operators an interoperable