From the NVIDIA edition of September 20, 2026
Emerald AI Hillsboro demo: full DSX Flex/Max-Q stack with utility dispatch scenarios
Emerald AI (with NVIDIA, Portland General Electric, EPRI) · AI data-center power flexibility / grid integration · United States (Hillsboro, Oregon; path to Manassas, VA)
- Publisher
- Emerald AI Blog
- Original publication
- March 16, 2026
- Source retrieved
- 2026-09-21
- Event date
- 2026-03-16
What happened
Emerald AI announced a Hillsboro, Oregon demonstration with NVIDIA, PGE, and EPRI integrating Emerald Conductor with the full NVIDIA DSX stack (including DSX Flex and DSX Max-Q) on Grace Blackwell Ultra clusters. PGE issued utility-style dispatch across scenarios: ~20% curtailment for freezing-rain/heating spikes; 25%+ curtailment across multi-day heat-dome re-enactments (6+ hour events over three days); a sustained 24-hour emergency response; and CAISO price-signal economic dispatch—while claiming priority workload performance was maximized under constraints. Positions the demo as the fifth live commercial-DC milestone toward a first commercial-scale DSX Flex deployment at NVIDIA’s planned 96 MW Vera Rubin AI Factory Research Center in Manassas, VA, and cites an aspirational “up to 100 GW” U.S. grid-capacity unlock.
Why it matters
Why now: prior Sep editions covered AEMA/Flex narratives and Lambda MaxLPS densification; this partner primary adds full-stack DSX Flex+Max-Q integration details and concrete curtailment percentages useful for SLED facilities/utility interconnection talks—without proving campus-scale results. Gov/Ed sites near constrained feeders may explore flex as interconnection strategy, not as guaranteed capacity.
Evidence and measured results
Dispatch scenarios with stated 20%, 25%+, and 24-hour responses; London precursor claimed >1/3 demand cut in under a minute on Blackwell Ultra; Nature Energy peer-reviewed Phoenix trial referenced; Manassas 96 MW commercial target named.
Limitations and uncertainty
Partner/vendor announcement; Hillsboro scale not MW-class in independent Volts commentary; 100 GW figure is system-level aspiration, not measured unlock; dated March 2026 but newly relevant to Sep Flex commercialization thread.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-09-21; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Interconnection delays and community pressure around inflexible AI loads.
- Stakeholders
- facilities
- utility relations
- CIO
- public affairs
- Discovery questions
- Is the site interconnection-constrained?
- Which workloads can flex for hours?
- What utility program exists locally?
- Value hypothesis
- Demonstrable flex may accelerate interconnection talks if curtailment depth/duration are contractually bounded.
- Potential engagement
- Joint utility tabletop + instrumented flex pilot on one hall.
- Unsupported claims
- 100 GW unlocked for the buyer
- 24-hour flex with zero priority impact at any scale
- Guaranteed faster interconnect
Pre-sales engineering
Role takeaway
- Fit
- Hyperscale/AI-factory sites with mixed training/inference; limited for always-on safety-critical only loads.
- Architecture and integration
- Emerald/DSX Flex signal path → workload scheduler/power caps → Max-Q/DSX telemetry loop.
- Prerequisites
- job priority taxonomy
- GPU/rack power telemetry
- utility dispatch interface
- SLO monitors
- Constraints
- Demo scale
- Partner authorship
- Scenario re-enactments not continuous production year
- Security
- Authenticate/authorize grid signals; protect orchestrator credentials.
- Proof of value
- Hit requested kW targets
- Priority job SLOs hold
- Audit trail of each event
Delivery
Role takeaway
- Work
- Integrate orchestrator, run scripted curtailment drills, capture power and SLO traces.
- Dependencies
- utility
- NVIDIA DSX stack access
- workload owners
- Ownership
- Facilities operations with platform SRE.
- Skills and adoption
- Ops need grid-event runbooks alongside GPU ops.
- Governance checkpoints
- Written interruptibility policy
- Community/regulatory review
- No capacity planning on 100 GW slogan
- Acceptance criteria
- Dispatch compliance logged
- Priority SLO breaches = fail
- Rollback to firm mode tested
- Risks
- Overstated scale
- Hidden priority degradation
- Signal spoofing
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?
Requires grid-signal ingestion, workload priority hierarchy, DSX Flex/Max-Q (or equivalent) power controls, and telemetry across GPUs/racks—not only static UPS/diesel planning.
Governance
Who approves, reviews and stays accountable for outcomes?
Utility program participation and community affordability claims need contractual clarity on who can interrupt which workloads and how SLOs are protected.
Security and privacy
What data, permissions and controls need testing?
Grid orchestration becomes a control-plane trust boundary; source does not detail adversarial robustness of dispatch channels.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Grid flex may protect residential rates (claimed) but can interrupt research/batch jobs—communicate impacts to academic users.
Procurement
What should contracts, pricing and exit terms secure?
May affect utility agreements and site selection RFPs; not a software SKU purchase alone.
Operating model
Which teams own the service once it runs?
Introduces shared ops between AI platform and utility demand-response processes.
Publication history
- 2026-09-20NVIDIA · Issue 158 resources
Stable resource ID: emerald-ai-hillsboro-dsx-flex-demo-20260316