NSF's $3.6M Convergence Grant Funds the Team, Not the Idea — Why Growing Convergence Research Rewards a Structure Most Proposals Get Wrong

There is a particular kind of research grant that punishes applicants for doing exactly what most of them think the program wants. NSF's Growing Convergence Research (GCR) program is one of them. On paper it looks like a large, generous award for ambitious interdisciplinary science — up to $1.2 million in Phase I and up to $2.4 million in Phase II, roughly $3.6 million over five years. In practice, it is one of the most conceptually demanding solicitations NSF runs, because it funds a thing — convergence — that is easy to claim and hard to demonstrate. The teams that win are the ones that understood the distinction before they wrote a word. The teams that lose usually assembled a strong multidisciplinary collaboration and assumed that was the same thing. It is not.

The current solicitation is NSF 24-527, with a recurring deadline on the second Monday in February — the next one falling February 8, 2027. NSF expects to make just 6 to 10 awards from a pool of approximately $16 million (pending appropriations). That is a small number of large awards, which tells you the review is unforgiving and the bar is set at a specific, well-defined concept. Understanding that concept is the entire game.

What NSF actually means by "convergence"

NSF defines convergence research through two non-negotiable elements, and both matter:

1. It is driven by a specific, compelling problem. The research must be organized around "a specific and compelling problem" arising from deep scientific questions or pressing societal needs — not around a discipline, a method, or a general theme. The problem comes first; the disciplines are recruited to serve it.

2. It requires deep integration across disciplines. This is the part applicants miss. NSF is explicit that convergence goes beyond people from different fields working in parallel. It requires that "knowledge, theories, methods, data, research communities and languages become increasingly intermingled." The disciplines do not just collaborate — they fuse into new frameworks, new vocabularies, and new ways of framing the problem that none of the contributing fields possessed on its own.

The word that separates convergence from ordinary interdisciplinary work is intermingled. A multidisciplinary project has a chemist, an economist, and a computer scientist each contributing their piece to a shared deliverable — three lanes running toward one finish line. A convergent project has those same three people building a shared conceptual framework in which the boundaries between chemistry, economics, and computation stop being meaningful for the problem at hand. NSF is funding the second thing. If your proposal describes the first thing in convergent language, experienced reviewers will see through it immediately, because distinguishing genuine convergence from repackaged collaboration is precisely what they are asked to do.

The two-phase gate is a filter, not a formality

GCR's structure is a two-phase design, and the gate between them is the program's central mechanism:

• Phase I (Years 1–2): up to $1,200,000 to establish the convergent team, build the shared intellectual infrastructure, and generate early evidence that the integration is real and productive.
• Phase II (Years 3–5): up to $2,400,000 in continuation funding — but only for teams that clear an "exceptional progress" review. Phase II is not automatic. It is earned.

This structure changes how you should think about the Phase I proposal. You are not proposing five years of research and asking NSF to fund the first two. You are proposing to prove that convergence can happen on your chosen problem, with your chosen team, within two years — and to produce the evidence that justifies the much larger Phase II investment. The Phase I deliverable is not a set of research findings so much as a demonstration that the team has genuinely intermingled and is generating frameworks neither discipline could have produced alone.

Teams that treat Phase I as "the first slice of a big project" tend to under-invest in the integration work — the shared workshops, the common data infrastructure, the deliberate cross-training that makes intermingling real — and then arrive at the exceptional-progress review with strong disciplinary results but weak evidence of convergence. That is the profile that does not advance to Phase II. The money is structured to reward teams that front-load the hard, unglamorous work of building a genuinely fused research community.

Team architecture is the proposal

Because convergence is fundamentally about how a team is built and how it works, the team architecture is the intellectual contribution of a GCR proposal — not a staffing appendix. Several structural realities follow:

Assemble the team around the problem, not the other way around. The strongest GCR proposals can point to a specific problem that no existing discipline can solve alone and then show why this particular combination of fields is necessary and sufficient to address it. If a reviewer can imagine solving the problem with fewer disciplines, or with a conventional collaboration, the convergence justification collapses.

Show the integration mechanism, concretely. Vague promises of collaboration are the single most common weakness. Winning proposals specify how intermingling will happen: shared physical or virtual space, joint appointments or embedded personnel, common data platforms, deliberate cross-disciplinary training, regular structured convening, and a plan for developing the shared vocabulary that convergence requires. NSF wants to see the machinery of integration, not the aspiration.

Include the right non-academic stakeholders from inception. NSF's framing emphasizes assembling "intellectually diverse researchers and stakeholders" from the start. For problems with societal dimensions, that can mean community organizations, industry, or government partners woven into the team's design from day one — not consulted at the end.

Note also the eligibility guardrail: a researcher may be PI or co-PI on only one GCR proposal at a time across active and pending projects, and violations trigger automatic desk rejection of the most recent submission. That rule forces investigators to commit to a single, serious convergent effort rather than spreading thin — another signal that NSF is buying depth of integration, not breadth of activity.

Why the timeline is an advantage, not a delay

The February 8, 2027 deadline is far enough out that it reads, at first glance, like a program to defer thinking about. That is exactly backwards. Convergence cannot be manufactured in a grant-writing sprint. A genuinely intermingled team is one that has already begun to build shared language and shared framing before the proposal is submitted — because reviewers can tell the difference between a team that has actually started working together and one assembled on paper for the application. The long runway to February 2027 is not slack; it is the minimum viable time to do the pre-proposal integration work that makes a GCR application credible.

The practical move for any research team eyeing GCR is to start convening now: identify the compelling problem, recruit the minimum necessary set of disciplines, and begin the actual work of building shared frameworks. By the time the proposal is written, the convergence should already be visible in early joint work, shared drafts, or a co-developed conceptual model. That evidence — proof that the intermingling is real and not promised — is what separates the 6 to 10 funded teams from the field.

The bigger picture

GCR sits inside NSF's larger, well-funded FY2026 research portfolio, and it represents a specific bet: that the hardest and most consequential problems now live in the spaces between disciplines, and that funding the formation of convergent teams is a distinct and worthwhile thing to do. For research institutions, the strategic insight is that GCR rewards a capability — the ability to build and sustain genuinely fused teams — that is rare, durable, and applicable well beyond this one program. The grant funds the team. Build the team, and the proposal writes itself.

To find NSF research solicitations and convergence-relevant programs matched to your institution and field, the Granted grant database tracks live federal research funding across every directorate.