Sparxity says Jülich deploys sparse computing rack

Sparxity, formerly SpiNNcloud, said it has deployed a sparse-native computing rack system at the Jülich Supercomputing Centre in Germany.

The company said the multi-million-dollar installation will be used for sparse machine learning models and large-scale high-performance computing simulations. It said the system, named JUBIS, is event-based and computes only when activity occurs.

Sparxity said JUBIS is designed to deliver 10 times faster sparse solver execution, 18 times lower energy consumption for event-based machine-learning workloads and 20 times shorter time to an acceptable solution in optimisation problems.

Dr. Guido Trensch of the Jülich Supercomputing Centre said sparse machine learning was an opportunity to improve AI efficiency and that adopting Sparxity opened new research opportunities in neuroscience.

The company also announced its rebrand to Sparxity, saying the new name supports its focus on sparse-native computing for AI and high-performance computing. Sparxity said the system is supported by the European Regional Development Fund and the state of North Rhine-Westphalia.