Adam J Morgan of NoMIS Power on Advancing High-Voltage SiC Technologies for Next-Generation HVDC Infrastructure

1. How does NoMIS Power's participation in the ARPA-E DC-GRIDS programme strengthen the company's strategic positioning within the rapidly evolving HVDC and grid infrastructure market?

Participation in DC-GRIDS positions NoMIS at the heart of one of the most strategically significant power-electronics initiatives the U.S. government is funding right now. ARPA-E's mandate is transformational, not incremental, and DC-GRIDS specifically targets HVDC architectures capable of substantially expanding U.S. transmission capacity. For a focused medium- and high-voltage SiC company like NoMIS, being selected as the device-and-packaging partner on a flagship project of this kind is a strong external signal — to utilities, OEMs, primes, and the federal customer — that our 3.3 kV portfolio is mature enough to anchor a next-generation submodule architecture.

Strategically, DC-GRIDS does three things for us at once. First, it accelerates our 3.3 kV technology roadmap by pulling our forthcoming devices into a real, demonstrable HVDC submodule application with a defined performance bar. Second, it embeds us in a consortium with the leading academic, national-lab, OEM, and utility voices in HVDC — Michigan State, NREL, EPRI, GE Grid Solutions, Salt River Project, and Minnesota Power — which is a vantage point that's almost impossible to manufacture artificially. Third, because the NPC-PEBB architecture is being developed as vendor-agnostic and plug-and-play, the work we do here is portable to other valve designs and other consortia — meaning the strategic dividend extends well beyond a single project.

It also reinforces a positioning choice we've been deliberate about. The big horizontal SiC players are concentrated on automotive volume at 650 V to 1.2 kV. NoMIS has chosen to specialize in medium- and high-voltage SiC, where design requirements are harder, the customer base is more strategic, and integrated capability matters more than wafer scale. We deliberately combine disciplines that competitors typically split across multiple organizations — SiC device design, advanced packaging and module development, and electrical screening and test — supported by a custom co-development model that lets us tailor devices, packages, modules, and test methodology to specific application requirements. That integrated stack is what HVDC submodules and other critical grid infrastructure actually demand from a solid-state solution, and it is the capability set DC-GRIDS validates at exactly the moment the grid-infrastructure opportunity is becoming a top-tier industrial priority in the U.S. and globally.

2. With increasing demand for grid modernisation, data centre power delivery, and offshore wind integration, how significant is the commercial opportunity for advanced SiC-based MT-HVDC converter technologies over the coming years?

The commercial opportunity is substantial, and we believe the medium- and high-voltage segment of SiC is one of the most under-discussed parts of the market story. The bulk of public market commentary on SiC has been about EV traction inverters at 800 V. That's a real and growing market — and industry analysts now project the overall SiC power-semiconductor market roughly tripling between 2025 and 2030 — but the MV/HV slice on top of that is being pulled by three demand vectors that are individually large and increasingly converging.

The first is grid modernization and MT-HVDC build-out. The U.S. transmission system needs a step-change in capacity to move power between regions, interconnect new generation, and harden against extreme weather. ARPA-E's framing on DC-GRIDS — that vendor-agnostic, plug-and-play valve technology is needed to unlock that build-out — captures the issue precisely. Every MT-HVDC valve in the new architectures will be built from submodules, and SiC-based submodules at the 3.3 kV and above class offer the efficiency, density, and reliability advantages that incumbent Si IGBT approaches can't match.

The second is hyperscale data center power delivery. The data center sector is now driving multi-GW additions of new load in single sites and corridors, and operators are moving toward higher-voltage DC distribution within and into the campus to reduce conversion stages, copper consumption, and footprint. MT-HVDC is the leading architecture for high-capacity delivery to data centre clusters from distant generation. SiC at 3.3 kV is the right device class for the converter stages that make hyperscale delivery economic.

The third is offshore wind interconnection and other long-distance renewable links. Offshore wind is essentially an MT-HVDC market by physics — distances are too long for AC. As offshore wind scales, the converter content per gigawatt is significant and increasingly favors SiC at medium- and high-voltage for the submodules inside the valves.

Stacking those three vectors, MT-HVDC SiC is one of the most attractive medium-term opportunities in the entire power-electronics space — characterised by high strategic value per converter, long deployment lifetimes, and a customer base that rewards focused MV/HV specialists.

3. NoMIS Power's 3.3 kV SiC MOSFET portfolio plays a central role in the NPC-PEBB architecture. From a business perspective, how does this project help validate and accelerate adoption of the company's high-voltage SiC technology roadmap?

From a business perspective, this project does for our 3.3 kV roadmap what flagship reference designs have historically done for new device classes in power electronics: it gives engineers, program managers, and prime contractors a concrete, credible system in which to see the technology working at the relevant voltage, current, and reliability levels.

The NPC-PEBB submodules being developed here are rated 6.6 kV / 2.5 kA — built directly from our 3.3 kV SiC MOSFETs and power modules. The roadmap that feeds this work is well-defined and progressive: the previously released NoMIS N3PT080MP330 (3.3 kV, 80 mΩ, 34 A) anchored our entry into the medium-voltage class; the 50 mΩ and 25 mΩ 3.3 kV MOSFETs that will complete the suite are the lead devices for HVDC submodule duty. The 25 mΩ device in particular is purpose-built for this application — the lowest possible on-resistance translates directly into reduced conduction losses, higher converter efficiency, and improved thermal headroom at MT-HVDC valve current levels. Having that device land into a consortium that will demonstrate it inside a real NPC-PEBB submodule is the kind of validation a roadmap engineer wants to see.

Three business outcomes follow. First, technical credibility. When a future utility, OEM, or system integrator asks whether NoMIS' 3.3 kV portfolio is HVDC-ready, we will be able to point to a demonstrated NPC-PEBB submodule built from our devices, characterised by us and tested in a multi-organisation consortium that includes NREL, EPRI, and GE Grid Solutions. That is a much stronger answer than a datasheet.

Second, design-in acceleration. The work covers electrical testing, screening, and performance characterisation of our SiC devices and power modules before NPC-PEBB assembly. The data we generate becomes reusable for any future customer evaluating these devices for medium- or high-voltage submodule applications, compressing what would otherwise be 12-to-24-month evaluation cycles. It also reinforces a structural advantage in our business model: because NoMIS combines SiC device design, advanced packaging, power module development, and electrical testing in-house, we can co-develop custom devices, packages, modules, and test methodologies tailored to specific application requirements. For customers with non-standard HVDC submodules, MV converters, or system-level needs, that single-partner path from chip to package to module to screening is a meaningful differentiator.

Third, supply-side optionality. Because the architecture is explicitly vendor-agnostic and plug-and-play, and because NoMIS' 3.3 kV portfolio is also available as supply to other DC-GRIDS teams and the broader medium- and high-voltage power-electronics developer base, we expect this project to seed engagements well beyond the lead consortium. It is, in effect, a force multiplier on the commercial pipeline for the entire 3.3 kV roadmap — and a stepping stone toward higher-voltage classes that follow on our portfolio plan.

4. The project brings together major industry and research organisations, including Michigan State University, the National Renewable Energy Laboratory, GE Grid Solutions, and EPRI. How important are collaborative industry partnerships in advancing next-generation power semiconductor and grid technologies?

They are essential. Power semiconductor innovation, particularly at medium- and high-voltage, is no longer a single-company exercise. The device, the module, the gate driver, the submodule, the valve, the converter, and the system around it must co-evolve because design trade-offs at one layer cascade up and down the stack. No single organisation has the full set of capabilities or the customer perspective needed to make that work. Consortia have therefore become the unit of innovation for this part of the industry. We have seen this pattern play out in adjacent semiconductor industries — including within the Albany NanoTech Complex, where NoMIS is headquartered.

In this consortium, every partner brings a distinct vantage point that NoMIS could not replicate internally. Michigan State University, under Dr. Omid Beik, brings the academic depth to multilevel converter architecture and control. NREL brings the national-lab capability on integration, testing, and grid-side characterisation. EPRI brings the utility R&D voice — what utilities actually need to see before they will deploy a new valve technology at scale. GE Grid Solutions brings the OEM perspective on what is manufacturable, qualifiable, and serviceable in a real HVDC converter station. Salt River Project and Minnesota Power bring the end-user view — the operational realities of running this equipment. OPAL-RT brings real-time simulation capability that links it all together. And NoMIS brings the SiC technology know-how across device, packaging, power module, and electrical screening and testing.

That last point is something we are very deliberate about. NoMIS combines four disciplines in-house — SiC device design, advanced packaging, power module development, and electrical screening and test — and offers a custom co-development model that lets us tailor devices, packages, modules, and test methodology to a partner's specific architecture rather than forcing the partner to design around a fixed catalogue. For a consortium working on a novel submodule architecture like the NPC-PEBB, that integrated capability is exactly what is needed at the SiC layer. We are not trying to be a horizontal supplier doing everything from automotive volume to the grid. By staying focused, we make a far better partner for consortia like this one, where the other organisations need a SiC partner who understands the device-and-package layer in the depth required by HVDC submodule duty.

Across NoMIS' history, we have consistently engaged with the U.S. federal R&D system — ARPA-E, the Air Force Research Laboratory, the Office of Naval Research, the Army Research Laboratory, NSF, and DOE — and with leading universities and national labs through programs like PowerAmerica and CASPER. DC-GRIDS is the natural progression of that work into a larger and more visible commercial application area.

5. The NPC-PEBB architecture is expected to deliver higher efficiency, improved power density, and enhanced reliability compared to conventional silicon IGBT-based systems. How critical are these performance gains in addressing future energy infrastructure and hyperscale power requirements?

These performance gains are not optional refinements. They are the prerequisites for delivering the next generation of energy infrastructure at the scale and pace it now needs to be built.

The NPC-PEBB approach delivers a clear step-change over conventional Si IGBT-based half-bridge submodules: a 3-level 6.6 kV output versus 2-level 4.5 kV; full DC fault current blocking; a 60% reduction in submodule capacitor size enabled by an advanced multilevel space vector modulation strategy; and improved efficiency, power density, and reliability across the valve and converter. Each of those translates into a concrete benefit for the infrastructure customer.

Take efficiency first. At hyperscale data-centre power levels — multi-hundred-megawatt to multi-gigawatt sites — every percentage point of converter efficiency turns into significant continuous power, cooling load, and electricity cost. SiC-based submodules at 3.3 kV reduce switching and conduction losses substantially versus Si IGBTs, and the multilevel architecture further reduces filtering and modulation losses. Across a single 1 GW MT-HVDC corridor, the efficiency dividend over a project lifetime is consequential.

Power density and footprint matter just as much. Land near major load centres — especially near data centre campuses — is constrained and expensive. The combination of higher-voltage submodule synthesis, reduced submodule count per valve, and smaller capacitor sizes lets a given converter station deliver more power in less space. That changes the siting calculus for new HVDC infrastructure.

Reliability is the third and arguably most strategic axis. Capacitors are one of the leading failure-rate components in HVDC submodules; the 60% capacitor-size reduction enabled by the multilevel space vector modulation strategy materially improves submodule, valve, and converter reliability. Full DC fault current blocking adds another resilience axis that conventional half-bridges cannot deliver without external solutions. For a utility or hyperscale operator, the move from "the converter is the limiting reliability factor" to "the converter is no longer the limiting factor" is the kind of step-change that reframes how they plan, deploy, and operate the asset.

Underlying all three is what SiC brings to medium- and high-voltage power electronics generally: a much better efficiency-ruggedness trade-off than legacy silicon, enabled by NoMIS' thicker-than-industry-standard gate oxide, optimised unit-cell structures, rigorous packaging discipline, and a design-for-test approach that includes aggressive device screening and binning. Those device-level fundamentals make the system-level NPC-PEBB gains possible.

6. Looking ahead, how does NoMIS Power see the role of Silicon Carbide technologies evolving across HVDC transmission, renewable energy integration, industrial electrification, and high-capacity data centre infrastructure?

Our view is that SiC will move from being an emerging technology in medium- and high-voltage power conversion to being the default — and that the transition will happen faster than most observers currently expect, because the demand-side pressures are now uncommonly synchronised.

In HVDC transmission, the trajectory is straightforward. As MT-HVDC architectures replace point-to-point legacy designs, and as new corridors are built to connect renewables, data centre load, and inter-region capacity, the valve content per gigawatt grows. SiC at 3.3 kV and above will increasingly be the device of choice inside those valves because it enables higher voltage synthesis per submodule, lower losses, smaller capacitors, and better fault response. Over time, we expect to see SiC at progressively higher voltages — 6.5 kV, 10 kV, 15 kV and beyond — opening up architectures that are not practical with silicon at all. NoMIS' portfolio extends to 20 kV, and the company has invested in MV/HV SiC device development across that full range under federal R&D programs. That higher-voltage trajectory is one we intend to lead from — by continuing to build on our SiC MOSFET offerings and, at the upper end of the voltage spectrum, by incorporating SiC IGBTs as well.

In renewable energy integration, the story is similar but earlier in its arc. Utility-scale BESS, high-power solar inverters, renewable converters, and MV/HV grid interfaces are all moving toward higher DC-link voltages to shrink balance-of-system cost. SiC at 1.7 kV and 3.3 kV — both of which are now in NoMIS' commercial portfolio — sit exactly at the device classes where this transition is happening. We expect medium-voltage SiC adoption in renewables to accelerate sharply over the next three to five years.

In industrial electrification — particularly medium-voltage motor drives and large-format VFDs — SiC enables more compact, more efficient drives at higher switching frequencies, which both reduces cooling load and increases the addressable applications. As industrial heat, mining, and large-scale process electrification scale, this is a structural growth driver. DC solid-state circuit breakers in MV DC networks are another adjacent application where SiC is a key enabler.

In high-capacity data centre infrastructure, the dynamics are the most aggressive of any sector right now. Hyperscale buildouts are pulling forward both higher-voltage DC distribution within the campus and MT-HVDC delivery to the campus from distant generation. SiC sits in the middle of that stack. The need for converter efficiency, density, and reliability at gigawatt-class campuses creates a strong pull for exactly the kinds of medium- and high-voltage SiC solutions NoMIS specialises in.

The throughline across all four sectors is the same: the world is asking power-electronics suppliers to move more power, more efficiently and reliably, in less space, and at higher voltages than legacy silicon was ever designed for. SiC is the answer. NoMIS' role is to be the focused medium- and high-voltage SiC partner of choice — bringing rugged, reliable, customizable devices, modules, and packaging to the customers and consortia building this next generation of infrastructure. DC-GRIDS is an important step on that path, and we expect to be doing a lot more of this kind of work in the years ahead.

Meet NoMIS Power at PCIM Expo 2026. NoMIS will be exhibiting at PCIM Expo 2026 in Nuremberg, Germany (9–11 June 2026), Hall 4, Stand 301, with a focus on its 3.3 kV and higher-voltage SiC MOSFET portfolio, power modules, and custom device, package, and module co-development for medium- and high-voltage power electronics customers. Visitors are encouraged to schedule meetings in advance via together@NoMISPower.com.