Alibaba's DAMO Academy has introduced the MAOSS model, which analyzes standard CT scans to help detect fatty liver disease in its early stages.
AI: Events
ZenDNN 5.2: How AMD Accelerated Language Model Performance on Standard Server CPUs
Infrastructure
AMD has released ZenDNN 5.2 – an update that significantly speeds up the execution of language models on EPYC server processors without the need for GPUs.
AI: Events
Managing Servers with Words: How Red Hat Taught AI to Understand Admin Tasks
Infrastructure
Red Hat has introduced an MCP server for Satellite, allowing system administrators to manage their Linux infrastructure using simple text queries.
Cursor has shared how its AI model evaluation system works – a hybrid approach that combines real-world scenarios and lab tests.
AI: Events
How Kimi Runs Hundreds of Thousands of AI Agents Simultaneously: A Look at the Infrastructure
Infrastructure
Exploring how Kimi created a scalable and secure environment for AI agents on the Alibaba Cloud platform.
AI: Events
SQL as a Language for 'Talking' with AI: What the Hologres and Model Studio Integration Offers
Products
Alibaba Cloud has integrated its Hologres database with the Model Studio platform, allowing data engineers to call language models directly from their SQL queries.
SGLang added support for the NVIDIA Nemotron 3 Super model on the day of its release, simplifying the creation of multi-agent systems based on efficient language models.
AI: Events
Speech Recognition in Noise: Why Systems Perform Well in Tests but Fail in the Real World
Development
We explore why speech recognition systems perform well in tests but struggle in real-world conditions with background noise.
Federated AI allows models to be trained where the data resides – without centralization. We examine how this works in practice within large organizations.