DIMAGGI AI
Pinned Loading
Repositories
- usable-compute Public
Turning AI infrastructure investment into usable compute: a portfolio covering scheduling, networking, reliability, power, cooling, placement, and governed operations.
- ontology-debt Public archive
Ontology Debt: audit an LLM's world-model against declared commitments — and track the debt. The measure leg of declare → measure → enforce.
- scheduler-vs-more-gpus Public
Turning scattered GPUs into working AI clusters: capacity accounting, an allocation-policy simulator, control-plane case studies, and a TPU parallel hypothesis. Slurm, Kubernetes, multi-cloud. Reproducible.
- reliability-economics Public
Which GPU-cluster recovery policy wins in which failure regime? A reproducible simulator (ETTR/MTTF/MTTR + dollars) and a policy phase diagram, validated against Meta's published reliability numbers. Spares vs elastic vs checkpointing.
- cooling-pue-ladder Public
The cooling ladder, executable: density gates, PUE-to-megawatts capacity math, ride-through seconds, and the two-fleets explanation of the PUE plateau
- slice-packer-torus Public
The allocation layer of a torus pod: which rectangles a request may take, what a dead chip costs, when isolation and wraparound contradict, and who may reconstitute a slice
- optical-circuit-intent Public
What to ask an optical plant for, and what it costs: circuit intents (request, hold-until, release, failover-to), retune legality windows, and OCS port packing
- span-contract Public
An admission contract for AI training jobs that cross a data hall: local / span / shrink / move / deny / escalate, machine-checked
People
This organization has no public members. You must be a member to see who’s a part of this organization.
Top languages
Loading…
Most used topics
Loading…