From the NVIDIA edition of September 19, 2026
SemiAnalysis InferenceX: early Rubin AgentX shows large tok/s/MW gains; MaxLPS not isolated
SemiAnalysis / InferenceX · Independent semiconductor / inference analysis · Global (benchmark fleet)
- Publisher
- InferenceX by SemiAnalysis
- Original publication
- September 14, 2026
- Source retrieved
- 2026-09-20
- Event date
- 2026-09-14
What happened
Sep 14, 2026 InferenceX port reports first verified AgentX agentic-inference results for Vera Rubin NVL72 on early pre-release software: large TCO and tok/s/MW advantages vs GB300 depending on interactivity target and engine (examples include ~67x throughput per owning-TCO vs GB300 Dynamo TRTLLM at 170 TPS on a dense NVFP4 path; ~7.2x tok/s/MW vs GB300 SGLang at 150 TPS; engine label critical—at 170 TPS vs GB300 TRTLLM the MW multiplier is ~62.9x but vs SGLang ~5.56x). Authors argue GTC 3x per MW sandbagged vs measured up to ~7x. MaxLPS is described as first-class dynamic power shifting to fit more GPUs from real workload power profiles, with PowerX integration forthcoming—AgentX curves are not presented as isolated MaxLPS measurements.
Why it matters
Independent counterweight to vendor AgentX/MaxLPS marketing for SLED capacity planning: multipliers are SLA- and engine-scoped, early-software, and not MaxLPS-isolated. Limited direct public-sector deployment; relevance is procurement literacy for tokens/MW slides.
Evidence and measured results
AgentX real-world coding trajectories; owning vs 3-year rental TCO frames; tok/s/MW table across TPS targets; MaxLPS qualitative description; early TRTLLM Rubin software caveat; subscriber-only continuation noted.
Limitations and uncertainty
Public port of newsletter; early pre-release software; NVIDIA collaboration acknowledged; some fleet economics are modeled; MaxLPS not separately ablated.
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
- Agentic tokens/MW marketing numbers are being treated as portable planning constants.
- Stakeholders
- AI platform
- CFO
- facilities
- Discovery questions
- Which P90 TPS and engine are in the slide?
- Is MaxLPS ablated or assumed?
- Early-software or GA stack?
- Value hypothesis
- Pinned AgentX-style comparisons can justify Rubin TCO—if SLA/engine are matched.
- Potential engagement
- Matched-SLA InferenceX-style bakeoff on buyer corpus.
- Unsupported claims
- 67x everywhere
- MaxLPS measured in AgentX curves
- GTC 3x is the ceiling
Pre-sales engineering
Role takeaway
- Fit
- Teams optimizing agentic serving efficiency under power caps.
- Architecture and integration
- Engine choice + interactivity target + power telemetry; MaxLPS evaluated separately.
- Prerequisites
- power meters
- AgentX or equivalent trace replay
- engine pins
- Constraints
- Early Rubin software
- modeled TCO components
- public article incomplete vs subscriber
- Security
- N/A.
- Proof of value
- tok/s/MW at matched TPS
- Engine A/B logged
- MaxLPS separate test plan
Delivery
Role takeaway
- Work
- Stand up pinned benchmark jobs and review board for multiplier use in planning.
- Dependencies
- SemiAnalysis/InferenceX methods or internal equivalent
- facilities metering
- Ownership
- Inference platform lead.
- Skills and adoption
- Analysts must read engine labels and TPS axes.
- Governance checkpoints
- No unpinned multipliers in budget decks
- MaxLPS claims require separate PoC
- Early-software watermark
- Acceptance criteria
- Pinned result archive
- Sensitivity to engine documented
- Planning number approved
- Risks
- Software churn
- Wrong SLA target
- Conflating TCO model with cash 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?
Compare engines at matched P90 TPS; do not collapse MaxLPS densification into AgentX efficiency curves; expect software maturity to move frontiers.
Governance
Who approves, reviews and stays accountable for outcomes?
Require engine/SLA/corpus pins on any Rubin multiplier used in budget cases.
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?
Require pinned SLA/engine exhibits alongside any Rubin efficiency claims in proposals.
Operating model
Which teams own the service once it runs?
Capacity economics tracked in tokens per megawatt at stated interactivity, not peak FLOPS.
Publication history
- 2026-09-19NVIDIA · Issue 148 resources
Stable resource ID: semianalysis-inferencex-rubin-agentic-20260914