HPC

High Performance Compute (HPC) clusters often produce intermediate files as part of code execution and message passing is not always …

A neural compression framework built on top of JAX.

DisTRaC is a deployment tool that allows Ceph to run on a High-Performance Computing cluster using RAM when a Network File System is present. It’s designed to be used with a job scheduler such as Univa Grid Engine (UGE) to create a transient object store as and when required which lasts as long as the user intends.

DosNa is intended to be a python wrapper to distribute N-dimensional arrays over an Object Store server. The main goal of DosNa is to provide easy and friendly seamless interface to store and manage N-Dimensional datasets over a remote Cloud.