Unlocking Productivity with Fractal Parallel Computing: A Geometry-Inspired ML Booster

Fractal parallel processing, a productivity booster inspired by geometry

We may browse the list of recommendations while trying to buy something with an app. The machine is already learning about us. Machine learning (ML), as an emerging technology that is effective, has been widely adopted. Its applications range from various apps to supercomputing.

Dedicated ML machines are being developed in various sizes, but they are limited by their low productivity. The workload and costs are largely concentrated on their software stacks which must be developed or reworked ad hoc to support each scaled model.

Researchers from the Chinese Academy of Sciences have proposed a fractal computing model to solve this problem. Their research was published in Intelligent Computing, on September 5.