Home / Compute & infra
Compute & infrastructure
A pooled GPU cluster and cloud-credit pool for active members. Enough compute to test an idea — not enough to brute-force a problem, by design. Allocations scale with contribution level, but every active member gets a baseline and project leads always get more.
What's in the pool
| Resource | Quantity | Where | Notes |
|---|---|---|---|
| H100 80GB SXM | 8 | Helsinki + Singapore | Shared via Slurm; max 7-day reservations |
| A100 80GB | 16 | Cluster + 4 community donors | Idle-time priority for project leads |
| A100 40GB / L40s | ~30 | Cloud passthrough | Allocated against $ credit pool |
| Cloud credit | $420,000 | AWS, GCP, Lambda | Quarterly burndown |
| Open datasets | ~3 TB | S3 mirror, eu-north-1 | Free for all members |
| Eval graders | — | Self-hosted | Used by competitions; free to run for any project |
How allocation works
- Every active member at L4+ gets $50/month in cloud credit by default.
- Project leads on accepted projects get up to 1 H100-week + $500/month by default.
- Fellows get $25,000 in cluster + cloud across their fellowship.
- Special-request bursts (3+ GPU-weeks, or $5k+ in cloud) get reviewed by the compute committee. Decisions in 7 days.
- Quarterly burndown: unused allocation rolls 50% into the next quarter; the rest goes back to the pool.
The compute committee
Compute is allocated by a 3-member committee, rotating quarterly, drawn from L8+ members. They are paid for the work. Their goal is to maximise research output per GPU-hour across the lab — not to maximise individual fairness. They publish a quarterly report showing where compute went, what came out of it, and what they would do differently.
Donating compute
If you have spare GPU time at home or in a lab, alphabell can use it. We have a turnkey contributor stack — Docker container, sandboxed, opt-in workloads only, audit logs. Donors earn contribution points for every GPU-hour donated.