UAPI Grok Burst Lane Cost Risk Matrix 2026-06-22
Grok cost, auth, lifecycle, and risk matrix for cloud burst lanes and required evidence before enabling.
Promotion Metadata
- Source feedback id:
132 - Source feedback ids:
132 - Promotion request id:
63 - Feedback category:
beneficial - Feedback source:
[REDACTED:high-entropy] - Feedback created at:
2026-06-22T18:32:58.892566+00:00 - Target slug:
[REDACTED:high-entropy] - Review state:
review - Reviewer note: AI lane auto-promoted redacted feedback into internal catalog context.
Source Feedback Body
Feedback 132
- Product:
uapi - Category:
beneficial - Related slug:
[REDACTED:high-entropy] - Source:
[REDACTED:high-entropy] - Created at:
2026-06-22T18:32:58.892566+00:00
Grok Burst Lane Cost-Risk Matrix (2026-06-22)
ALIVE: true
Lane: Grok Worker D / UAPI burst provider comparison
Sources: agent-work-queue-v2-unblocked.md; local (grok-cloud-colab-burst-lane-review.md, grok-cloud-gpu-broker-contract-notes.md + priors); public: https://research.google.com/colaboratory/faq.html, https://colab.research.google.com/signup, https://docs.runpod.io/serverless/pricing, https://www.runpod.io/pricing.
Broker Context (from prior)
UAPI lease + artifact handoff only (pre-lease sink inputs; post outputs). Hard caps. Local/lab first. https://docs.runpod.io/serverless/overview
Cost/Auth/Lifecycle/Risk Matrix
| Aspect | Colab Notebook Burst | RunPod Serverless/On-Demand | Lab-hosted (NUC 140T/NPU, Cortex) |
|---|---|---|---|
| Cost (marginal) | $9.99/100CU; T4~1.5-2CU/h, A100~5-15. Unpredictable burn. | ~$0.00019-0.00116/s (A100 equiv ~$2.7/h util). No idle Flex. | $0 marginal; amortized hw+power. Cap by inventory. |
| Predictability | Low (fluctuate, exhaust term). | High (per-sec + $80/h limit). | Highest (fixed). |
| Auth | Google+billing/Drive. Re-init/lease. | API keys + vols + endpoints. | Internal op-mediated (no cloud keys). |
| Isolation | Paid config only. | Per-lane endpoints. | Highest (lab bounds). |
| Lifecycle | Ephem VM (idle+12-24h max). Notebook bias. | Scale0/Flex/active. Cold (FlashBoot mit). Timeouts. | Persistent/reserved. Re-boot post. |
| Startup/Term | Fast connect; high term risk (priority/quota). | Cold secs; controlled timeouts. | Near-0 or op; low ext risk. |
| Key Risks | Batch disallowed (FAQ); Drive quota; no SLA. | Cold cost; image stream; avail. | Limited cap; Conductor contention; gate dep. |
| Exposure/Lock | Drive+ephem high lock. | Docker portable; sink mit. | Lab sinks; lowest lock. |
Evidence Required (Pre-Enable)
- Isolated paid config/key (UAI rules; never shared).
- UAPI hard cost cap + accounting (report deltas; Colab soft insufficient).
- Artifact roundtrip (manifest pull + result/logs push).
- Synthetic: lease, hard/idle timeout enforce, partial flush, release.
- Health: pre-dispatch avail/GPU/store; ongoing via artifacts.
- Placement rules + Codex baseline (local/lab > burst on budget).
- Fail modes: quota, disconnect, overrun, partial, auth.
- Sanitized evidence + operator gate (after local routes).
- Broker contract + adapter reviewed.
All: artifact resume; dual timeouts; no provider state.
Assumptions: Artifact sinks durable; small inputs pre-warmed (cold fetch wastes GPU); multi-provider future. Blockers: No configs/keys; contracts absent; no synthetic runs; NUC/SN850X inventory separate. Next: Codex promote matrix + comparison; define adapter evidence checklist; operator gate for isolated config validation only.
Report: ALIVE: true | Artifact: [REDACTED:high-entropy].md | Cost/auth/lifecycle/risk matrix + evidence checklist for burst lanes. Cited public+local. Local only. <3800. COMPLETE. End.