From the NVIDIA edition of September 19, 2026
NVIDIA DSX MaxLPS overview: MaxP vs MaxQ vs MaxLPS and illustrative 1.35x tokens
NVIDIA · AI factory documentation · Global (vendor docs)
- Publisher
- NVIDIA DSX Documentation
- Original publication
- undated docs page (retrieved 2026-09-19)
- Source retrieved
- 2026-09-20
What happened
Official DSX MaxLPS overview defines MaxLPS as coordinating facilities/site design, Dynamic Power Software (DPS), and advanced perf/W techniques to maximize throughput per megawatt in a fixed site-power envelope, claiming up to 40% more GPUs in that envelope. It contrasts static MaxP/MaxQ GPU settings with factory-wide dynamic MaxLPS, and shows an illustrative Vera Rubin NVL72 inference scenario of 400 GPUs per 1 MW and 1.35x token throughput vs MaxP (PUE 1.1), while stating Vera Rubin testing remains in progress. Emphasizes planning for 45C cooling even when Day-1 training mixes offer less headroom than inference.
Why it matters
Clarifies product taxonomy for SLED facilities evaluating MaxLPS after Sep 18 Lambda PoC coverage: distinguishes marketing densification from design prerequisites (45C, lifecycle fill). Newly archived docs page; treat illustrative 1.35x/400 GPU/MW as planning graphics, not measured customer results.
Evidence and measured results
MaxP/MaxQ/MaxLPS mode comparison; up-to-40% more GPUs claim; 45C cooling callout; illustrative 1 MW scenario 400 GPUs / 1.35x tokens vs MaxP; note Vera Rubin testing in progress; stack diagram across facilities/hardware/infra software/platform software.
Limitations and uncertainty
Vendor docs; undated page; illustrative scenario; Rubin testing in progress disclaimer on the same page.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-09-20; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Buyers confuse static MaxQ settings with factory-wide MaxLPS densification claims.
- Stakeholders
- facilities
- platform
- procurement
- Discovery questions
- Is 45C DLC in the design?
- What PUE assumption are you using?
- Do you have a measured PoC beyond the overview graphic?
- Value hypothesis
- MaxLPS-ready design can raise lifecycle tokens/MW inside a fixed envelope if illustrative figures stay labeled as such.
- Potential engagement
- MaxLPS taxonomy workshop plus site illustrative-to-actual model.
- Unsupported claims
- 1.35x tokens guaranteed
- 400 GPUs/MW universal
- Rubin MaxLPS fully validated
Pre-sales engineering
Role takeaway
- Fit
- Design-phase AI factories; weak as sole evidence for production densification.
- Architecture and integration
- Facilities + DPS + workload perf/W stacked under fixed MW.
- Prerequisites
- cooling target
- power telemetry
- network sized to target GPU count
- Constraints
- Illustrative graphics
- Rubin testing in progress
- Day-1 mix dependency
- Security
- N/A in source.
- Proof of value
- Replace illustrative MW model with site meters
- Cooling inlet achieved
- DPS policy pilot
Delivery
Role takeaway
- Work
- Update design standards and pilot plan; keep overview as reference only.
- Dependencies
- MEP
- NVIDIA DSX SE
- platform
- Ownership
- Facilities architecture lead.
- Skills and adoption
- Teach MaxP/MaxQ/MaxLPS vocabulary to capital projects teams.
- Governance checkpoints
- Illustrative figures footnoted
- No budget lock on 40%
- 45C gate
- Acceptance criteria
- Checklist adopted
- Site model filed
- Pilot scheduled
- Risks
- Docs treated as measurement
- Training mix blocks fill
- Cooling retrofit cost
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Design power, cooling, and network for MaxLPS target GPU count even if Day-1 population is lower; DPS reclaim is necessary but not sufficient without facilities readiness.
Governance
Who approves, reviews and stays accountable for outcomes?
Label overview graphics as illustrative; require site pilots before budgetizing 40% or 1.35x.
Security and privacy
What data, permissions and controls need testing?
Not addressed.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
None stated.
Procurement
What should contracts, pricing and exit terms secure?
Reference docs for scope language; attach measured PoC exhibits to contracts.
Operating model
Which teams own the service once it runs?
Capacity planning becomes lifecycle fill under a fixed MW envelope.
Publication history
- 2026-09-19NVIDIA · Issue 148 resources
Stable resource ID: nvidia-dsx-maxlps-overview-docs-2026