Roadmap¶
Phase 0 · Live¶
HTCondor at MIT-WPU.
The first compute pool runs at the MIT World Peace University Pune research-computing group, scheduled by HTCondor — the same production scheduler used by the CERN LHC computing grid, the NASA HEC programme, and roughly 100+ universities and labs worldwide.
Researchers submit jobs to a real HTCondor pool with the queue behaviour, scheduling semantics, and resource matching they would get on a national supercomputer. The pool has uptime, monitoring, and an escalation path.
Phase 1 · In motion¶
Multi-university federation.
Two additional Indian universities are joining as partner sites. Each new partner contributes compute capacity and gets access to the full pool in return.
- Cross-site identity so a researcher at one partner can run a job that lands at any other partner.
- Hardened operator tooling for unattended day-to-day operation.
- Partner-onboarding playbook a single campus-side administrator can run.
Phase 2 · Planned¶
The full research-computing stack.
Once federation is stable, the grid becomes the delivery vehicle for a broader set of open-source research-computing tools — behind a single login, all permissively licensed.
- Web job submission. No SSH.
- Interactive notebooks and remote desktops on cluster hardware.
- Project tracking, group documentation, data versioning.
- Operator observability partners can self-audit.
Funding¶
Phase 0 is operated by the MIT-WPU research-computing group. Phase 1 funding is in active conversation with academic-partner sponsors and research-infrastructure investors. See contact.