Shaping the NSTC’s IP Policy – Part II

Drawing the inevitable conclusion from a set of givens

~Cover photo designed on Midjourney~

Who should own the innovations produced by the US’s historic new semiconductor research consortium?

The National Semiconductor Technology Center (NSTC) is an ambitious public-private partnership receiving $9 billion in funding from the CHIPS and Science Act of 2022. Intended to support breakthrough innovations across the chip industry, the NSTC will undoubtably generate valuable technology. Of great concern to policymakers, the consortium’s intellectual property (IP) policies will determine who can monetize on these innovations.

As we learned in Shaping the NSTC’s IP Policy – Part I, officials from the US Department of Commerce tasked with implementing the NSTC have limited discretion in designing IP policy. The experiences of consortia around the world demonstrate that funding sources, membership, and research agenda limit the world of viable IP models officials can choose from. Critically, selecting an IP model incompatible with the consortium’s fundamentals may jeopardize the whole endeavor.

This article applies lessons from Part I to identify what IP models suit the NSTC. Today, we will:

• Establish the NSTC’s likely funding structure, membership, and research agenda and

• Discuss how structural decisions for the NSTC necessitate certain IP models.

If after reading Part I and II you are sick of IP but still interested in semiconductor research content, check out Chip Capitols’ article on NSTC geography.

Inputs to the IP Policy Equation

Commerce Department officials have limited discretion in designing the NSTC’s IP policy because IP is downstream from decisions relating to the center’s funding, membership, and research agenda. This section draws on comments from industry, academia, and other actors in the semiconductor ecosystem to predict how officials will design these three fundamentals.

Funding

Public private partnerships like the NSTC can choose from many funding models, but some of these may undermine subsequent IP policies. A recently released report by the nonprofit MITRE Engenuity does an excellent job laying out a combination of funding sources that could make the NSTC financially sustainable in the long run. If not implemented with a proper IP policy, however, some of these revenue streams risk double-dipping on jointly developed technology in an unsustainable way.

Federal government funding from the CHIPS Act currently provides up to $8 billion to support the NSTC over five years. Whether this comprises a large or small share of the NSTC’s total budget will depend on how many other sources of revenue it brings in, and the relative burden taxpayers carry will impact IP decision making. Industry and third-party observers generally agree that federal funding should comprise less than a majority of the NSTC’s funding long-term.

State and local support comes with different IP policy considerations than federal CHIPS Act dollars. Whereas Congress’ goal in funding the NSTC was to bolster American innovation, local governments primarily offer incentives to attract jobs. National innovation benefits from a more open-access IP policy, but the NSTC’s IP access limitations are unlikely to change the number of jobs the center creates. This funding stream is thus unlikely to have a significant impact on the NSTC’s IP policy.

Membership fees are the entrance dues entities pay for access to the NSTC’s physical facilities, participation in its research programs, and access to its shared IP. MITRE and several large semiconductor companies have specifically proposed a similar tiered membership structure for the NSTC, where smaller participants could pay less for narrower cooperation arrangements. MITRE has also suggested allowing participants to earmark part of their dues to support the NSTC initiatives of greatest interest to them.

Contract work would involve companies ranging from startups to established firms paying for access to NSTC prototyping equipment that would be too expensive for any individual company to maintain on its own. Whether the NSTC can engage in contract work, however, will depend on the scheduled availability of physical equipment. Disproportionate funding from members will likely mean facilities must prioritize joint projects involving those stakeholders.

IP licensing would involve the NSTC offering non-members access to technology developed through consortium activities. Observers doubt that this could be an immediate revenue stream because many NSTC technologies will take several years to develop and the center will face pressure to license technology to members only.

Membership

The Department of Commerce made clear in its CHIPS Strategy memo that the NSTC will have a diverse membership including “startups, small firms, fabless semiconductor companies, in-house semiconductor designers, allied international entities, and universities,” as well as large chip manufacturers.

Given this diversity, the NSTC will almost certainly have different tiers of membership to accommodate participants with less IP or cash to offer. Established semiconductor companies have made it clear they expect lower paying participants to receive fewer benefits from the consortium. This will impact how broadly the NSTC can share its IP among members.

Also, because the federal government will be a major, if not the biggest, financial contributor to the NSTC, the Bayh-Dole Act will be relevant to the center’s IP policy. Specifically, Section 401.14(e)(1) of the Bayh-Dole Act entitles the U.S. government to “march-in” rights. These allow the government to license patents developed with federal funds to other parties if the existing patent holder is not making sufficient commercialization efforts. Though the U.S. government has never exercised its march-in rights, effective communication with the private sector (especially, startups) is critical to assure prospective members that the government will not suddenly appropriate jointly developed IP.

Agenda

The NSTC’s research agenda will be focused on great leaps forward in semiconductor technology, but it remains unclear whether these leaps will be born from full-stack plans or parallel technology siloes. Such organizational decisions will play a major role in the NSTC’s innovation output, they will not significantly differ in their impact on its IP policy.

Generally favored by broad coalitions including academia, a centrally planned, full-stack agenda will attempt to solve “breakthrough challenges” that no company could approach individually. The NSTC would organize its efforts around problems like datacenter energy consumption, with specific technology roadmaps for logic and memory chips serving to achieve the larger breakthrough challenge.

In contrast, major semiconductor manufacturers tend to prefer organization around specific technology verticals, namely memory, sensing, and advanced packaging. The technology roadmaps here may not be as integrated into a larger NSTC-wide goal, and individual companies will be more likely to spearhead (financially and logistically) the verticals they specialize in.

Whichever organizational approach the NSTC adopts, the center will produce significant amounts of both early-stage and late-stage IP. The NSTC’s IP policy will thus need to account for both.

Models and Tools

Policymakers must pick IP models that facilitate commercialization of the NSTC’s research, but some IP policies will simply not be compatible with the NSTC’s funding, membership, and research agenda.

An excellent 2021 paper by the Institute for Defense Analyses lays out the four most prominent IP models for PPPs. The next section will show why an approach branching the second and third models is most appropriate for the NSTC.

1. Public/Open Access: At the least restrictive end of the spectrum, consortia could make all IP open to the public by publishing results and not requiring licensing fees to implement findings. These are most common for PPPs focused on basic research.

2. Shared Limited to All Partners Only: Here, members have free access to any IP developed by the PPP through nonexclusive royalty-free licenses. Use may be limited, though, to “evaluative work” that does not solely benefit an individual firm’s priorities.

3. Shared Limited to R&D Collaborators: At the tightest end of the sharing spectrum, only members who participated in a specific technology’s development own the resulting IP. Access for other consortium members varies.

4. Exclusive Model: Some consortia license technology on an exclusive basis to third parties. These tend to occur in PPPs focusing on late-stage research, and such licenses are a significant source of income for those consortia.

Outputs from the IP Policy Equation

At first glance it may seem premature to predict the range of IP models available to the NSTC since uncertainty remains over its funding and agenda. (We can be certain that the NSTC’s membership will be diverse.) In fact, however, we already know enough to predict that the NSTC will need a hybrid shared limited IP model. Here, we will plug in the inputs (funding, membership, and research agenda) to show the narrow IP policy line Commerce Department officials will need to walk.

Funding

Though the exact make-up of the NSTC’s funding is in flux, this pillar calls for a hybrid shared limited model. This hybrid model exists somewhere on a spectrum where IP is available royalty-free to all NSTC members at one extreme and available only to NSTC members who participated in a particular technology’s development at the other extreme.

Federal Funding – Part I showed that consortia like Japan’s AIST or Taiwan’s ITRI, which are majority government funded, operate under more open-access IP models. This is partially because government stakeholders prioritize getting public benefits out of research consortia, and, absent equally large private sector contributions, there is less reason to limit IP rights to consortium members. Advocates across the board expect less than half of the NSTC’s budget to consist of federal funding, so there will be less pressure from the government to make innovations public.

Membership Fees – As we saw in Part I, consortia like Belgium’s Imec that rely heavily on membership fees tend to compensate their members by giving them shared access to consortium-generated IP, which is not open to the general public. In order to facilitate participation by startups, Imec also offers lower tiers of membership that accord with more limited access to jointly developed IP.

Since membership fees are likely to be the NSTC’s biggest counterbalance to federal funding, members will expect significant benefits that are not available to the public. Furthermore, under the widely supported proposal for a tiered membership system, higher paying members will expect greater benefits than those available to lower paying members. This funding stream pushes the NSTC toward a default of sharing IP internally with all NSTC members, with a mix of limitations for lower paying members (see next subsection on membership for more).

Contract Work & IP Licensing – Having contract work as a major revenue stream empowers consortia to fund the rest of their agenda more independently. That independence in turn allows consortia to own more of their own IP, which can then be monetized via exclusive licensing arrangements.

The ability to engage in significant contract work represents a chicken and egg problem, however. Consortia like Germany’s Fraunhofer that receive a vast majority of their annual budget from contract revenue are only able to engage in such high levels of contract work if their facilities are not already earmarked for member or government research activities. If the NSTC relies heavily on federal funds and membership fees, these stakeholders would be upset if the consortium engages in contract research to be licensed to third parties.

If the NSTC nonetheless attempts to license jointly developed technology to third parties, it risks backlash from fee-paying members. As mentioned in Part I, China’s National High-Speed Train Technology Center (NHSTTC), strictly limits its reliance on these licenses for fear of alienating members that contributed to the IP’s development. Significant contract work or third-party IP licensing would likely be incompatible with a funding model that also relies heavily on membership fees.

Membership

Given that the Commerce Department is committed to making the NSTC conducive to a diverse membership, the default policy of sharing IP with all members must come with caveats.

On a high level, the participation of large competing semiconductor manufacturers means even IP shared among all members must have limited uses. Both a major trade association and a coalition of academia and industry have endorsed the idea of “evaluative licenses,” where consortium members can only use jointly developed IP for pre-production assessments –not commercial use.

Going down the membership tier ladder to lower paying members, the NSTC will need different policies for managing smaller companies’ access to rights for purchased IP, foreground (jointly developed IP), and background IP (technology developed outside the NSTC). Access to purchased IP is the least likely to change for lower tier members. The NSTC will annually spend part of its budget on software and process kit licenses that companies and startups across the chip industry need to research new technology. Since there is no unique competitive advantage at stake, larger companies may be content with the NSTC providing startups and other lower tier members with partially subsidized access to purchased IP.

In contrast, foreground IP will need greater protections. Higher tier members will not want lower tier members to free-ride off the NSTC’s innovations, so access to at least some jointly developed IP will need to be limited to only members who contributed to the relevant project. A more dynamic alternative to strict limitations could be an internal market trading system where NSTC members could purchase access to IP as needed. Lastly, large companies will be most protective of their background IP. Top tier NSTC members will contribute background IP as part of their member fees to support key projects, so they will be skeptical of lower tier members being given access to this IP for non-joint research project.

Agenda

Whether the NSTC adopts a centrally planned, full-stack research agenda or an approach centered around specific technology verticals, it will produce both early-stage and late-stage innovations.

Early-stage technology is not immediately commercializable, and the NSTC projects that develop it will likely use resources from a broad swath of members. The IP generated from these efforts can thus be licensed under more inclusive conditions to all members of the consortium. (This is also the approach taken by many university-industry collaborations.)

Late-stage technology is much easier to commercialize, and companies specializing in relevant technology verticals (analog, memory, etc.) are likely to make disproportionately higher contributions to the development of these innovations. These larger companies will thus likely want late-stage IP be limited to members who contributed to its development.

Structural Decisions Are Handcuffs

After reading this article and (if you are a true fan) the Part I article, you could be forgiven for thinking this was simply an analysis of what IP policy the NSTC should adopt. Yes, it is interesting that the consortium will likely need to share IP among its members with some variation for different tiers of participants, but there is deeper point.

Commerce Department officials’ early structural decisions will limit the choices they can make later. In this case, we saw that IP policy choices are downstream from decisions regarding the NSTC’s funding, membership, and research agenda. In an earlier Chip Capitols article, we saw that whether the NSTC is geographically centralized or decentralized impacts the research agenda it can adopt. Across our international case studies, we also see this trend of structural decisions impacting downstream policy choices.

As US government officials solidify their plans for the CHIPS Act’s research programs, Chip Capitols will continue analyzing the NSTC and research consortia around the world for lessons to inform the implementation of this historic public private partnership.