NSF's IDSS Program Has a $30M Ceiling and a July 28 Deadline — But the One-Proposal Rule Decides Your Whole Team's Strategy.

Modern science runs on data infrastructure that almost no one outside a handful of facilities ever sees: the repositories, pipelines, and services that let a genomics consortium, a climate-modeling collaboration, or an AI research team actually store, move, and compute on data at national scale. When that infrastructure works, it's invisible. When it doesn't exist — or exists only as a fragile pilot on one campus — entire fields slow down waiting for it. The National Science Foundation's Integrated Data Systems & Services (IDSS) program, running in 2026 under solicitation NSF 26-509, is the mechanism NSF uses to fund that infrastructure into existence and keep it running.

IDSS is not a typical research grant, and treating it like one is the fastest way to lose. It puts an estimated $60 million behind operations-level national cyberinfrastructure for open, data-intensive, and AI-driven science, with individual awards ranging from a $500,000 floor to a $30,000,000 ceiling. The full-proposal deadline is July 28, 2026 for the relevant categories. And buried in the mechanics is a participation rule that reshapes how any large institution should assemble its teams. This is the deep dive on how IDSS is structured and what winning it actually requires.

Operations, Not Discovery

The first mental adjustment is the hardest for research faculty. Most NSF programs fund discovery — you propose to learn something new, and the deliverable is knowledge. IDSS funds operations — you propose to build and run a system that other researchers depend on, and the deliverable is uptime, reliability, and service.

The solicitation's own language is precise about this: it supports "operations-level national-scale cyberinfrastructure systems and services that broadly advance and facilitate open, data-intensive and artificial-intelligence-driven science and engineering research, innovation, and education." Read that as a mandate for production-quality infrastructure, not a prototype. Reviewers will judge you on whether your system can serve a national user community at operational quality — security, availability, support, sustainability — not on the novelty of a single research result.

The AI framing is central and current. As machine-learning-driven science generates and consumes data at a scale that breaks older systems, NSF is explicitly funding the data cyberinfrastructure that AI-driven research demands. If your proposed system enables large-scale AI or data-intensive workflows across institutions, you are aimed at the heart of what IDSS wants to fund in 2026.

The Three Categories — and the Money Behind Them

IDSS accepts proposals in three categories, and they represent fundamentally different stages of maturity:

• Category I — Build new. Development, deployment, and operation of novel national-scale integrated data systems and services, potentially interfacing with or leveraging existing capabilities. This is the greenfield lane: you are proposing to create national infrastructure that doesn't yet exist.

• Category II — Scale up. Transition of an established smaller-scale, regional, pilot, or prototype data system to national-scale production/operational quality, including enhancing or expanding existing national operational systems. This is the "we have proof it works locally, fund us to make it national" lane.

• Category III — a distinct track with its own scope, sharing the July 28 full-proposal deadline alongside Category I.

The funding envelope is wide: with an estimated $60 million available, an award floor of $500,000 and a ceiling of $30 million, IDSS spans everything from a focused service to a flagship national facility. That range is a strategic signal. A $30 million Category I award is a multi-year, institution-defining commitment that NSF will scrutinize for sustainability and governance. A $500,000 award is a far more contained bet. Matching your ambition — and your institution's genuine operational capacity — to the right point on that spectrum is half the battle. Proposing a $30 million national system when you have run a single-campus pilot invites exactly the credibility question that sinks reviews.

The One-Proposal Rule Is a Team-Design Constraint

Here is the mechanic that most affects large institutions, and the one that has to be settled before anyone starts writing: an individual may participate as PI, co-PI, or other Senior Personnel on at most one proposal across Categories I and II for each solicitation deadline.

Read that carefully. It is not one proposal per category — it is one proposal across both categories combined. If Professor Chen is senior personnel on a Category I "build new" proposal, she cannot also appear on a Category II "scale up" proposal from the same or any other institution for that deadline. Her name is spent.

For a research university with multiple groups eyeing IDSS, this forces an internal coordination problem that is easy to discover too late. Your most senior, most credible infrastructure people — the ones whose names lend a proposal weight — are a scarce, non-duplicable resource across your entire IDSS portfolio. If two teams both want the same star co-PI, they cannot both have her, and someone has to choose which proposal she strengthens. Institutions that sort this out in June write clean, well-staffed proposals. Institutions that discover it in late July end up with two half-strength submissions or a last-minute scramble to re-paper their senior personnel.

The practical move: before drafting, map every intended PI, co-PI, and senior person against every IDSS proposal your institution might submit, and resolve conflicts deliberately. Treat senior-personnel allocation as a portfolio decision, not a per-team one.

Who Should Actually Apply

IDSS is not for individual investigators pursuing a research question. It is for teams and institutions — often multi-organizational — that can credibly stand up and sustain national infrastructure. The strongest applicants tend to share a few traits: a demonstrated track record of operating systems (not just building them), a real and identifiable national user community waiting for the service, and a serious plan for security, governance, and long-term sustainability beyond the award period.

That sustainability question deserves emphasis because it is where operations-focused proposals most often fail. NSF is wary of funding infrastructure that goes dark the moment the grant ends. A proposal that treats sustainability as an afterthought — "we'll seek follow-on funding" — reads as a liability. A proposal that names its user base, its governance model, and its path to durability reads as an investment. For Category II especially, the fact that you have already run a working pilot is your single best evidence that you can operate at scale; lead with it.

The Deadline and the Path In

The full-proposal deadline for the relevant categories is July 28, 2026. That is the immovable date. But the work that determines whether you can hit it credibly starts well before — assembling a multi-institution team, resolving the one-proposal senior-personnel conflicts, documenting your operational track record, and building the sustainability and governance case that operations-focused reviewers demand.

If you are contemplating IDSS, the sequencing is clear. First, decide which category genuinely fits your maturity: are you building something new (I), scaling a proven pilot to national production (II), or pursuing the distinct Category III scope? Second, lock your senior-personnel roster across all categories so no name is double-committed. Third, build the proposal around operations — reliability, security, sustainability, and a named national community — rather than around a research result. IDSS money is substantial, ranging up to $30 million, but it flows to teams that can prove they will keep the lights on for the science that depends on them.

Following NSF cyberinfrastructure and AI-infrastructure funding? See Granted News for the latest NSF solicitation announcements, and search Granted for the full slate of NSF programs with July and August 2026 deadlines.