Intellectual hub of the topic
AI: Events
Technical context • Infrastructure
AMD has unveiled a method for partitioning a single GPU into isolated domains to run different models simultaneously – with no compromise on security or performance.
AI: Events
Technical context • Development
Mistral AI engineers shared how they tracked down a memory leak in the popular vLLM language model serving system and what challenges they faced.
AMD has unveiled TileLang – a tool that simplifies writing optimized GPU operators and lowers the barrier to entry for ROCm development.
AMD is set to build the Genesis supercomputer for the U.S. Department of Energy – a machine powered by EPYC processors and Instinct MI300A accelerators, designed for scientific tasks utilizing AI.
Lab
Electrical Engineering & System Sciences
I'm explaining how, using mathematical transformations, we sped up encrypted control systems tenfold – without compromising security.
AMD has published benchmark results for its new Instinct MI355X GPU in neural network inference tasks, covering both single-node and distributed system performance.
Radio telescopes generate petabytes of data, but what if most of it can be predicted mathematically? This is the story of how to compress cosmic signals losslessly.
Lab
Electrical Engineering & System Sciences
We break down how a lightweight neural network identifies modulation types in OFDM systems, saving computing power without sacrificing accuracy – technology built for the real world.
Lab
Electrical Engineering & System Sciences
A new antenna architecture for 6G uses simple geometry instead of thousands of phase shifters – cutting costs by 15x while maintaining connection efficiency.
Don’t miss a single experiment!
Subscribe to our Telegram channel —
we regularly post announcements of new books, articles, and interviews.