Job search
DOE SCGSR for Computing PhDs: Match the Thesis, Lab Scientist, and Research Period
Match a defined computing thesis to a current DOE research priority, qualified laboratory scientist, continuous research period, and supplemental funding scope.
DOE SCGSR supports a defined part of an existing doctoral thesis at a participating national laboratory or facility. For a computing PhD candidate, the application decision starts with three connected questions: what thesis problem needs the laboratory, which scientist can collaborate on it, and whether a continuous research period fits the degree plan.
It is supplemental research support, not general funding for starting a PhD or a promise of a laboratory job. This guide uses official sources checked October 23, 2026, with completed fictional research and calendar examples. If employment is your immediate priority, explore recent technology roles on LandOffer as a separate search.

Confirm doctoral candidacy before researching a host
The current SCGSR eligibility page requires full-time enrollment in a qualified PhD program, candidacy already obtained at application, a defined thesis project, and a primary thesis advisor. Computer and computational sciences are eligible fields; the proposed thesis research must still align with the current priority areas.
Candidacy proof needs particular attention. DOE specifies an authorized, signed institutional letter certifying completed required coursework and a committee-approved thesis topic, including its approval date. It identifies acceptable signatories such as the registrar or department chair, head, or graduate-studies director; an administrative staff signature is not sufficient. Do not substitute “qualifying exam scheduled” for candidacy already completed.
The program also requires U.S. citizenship or lawful permanent residence at application and a minimum age of 18. Lab access remains a separate approval. A full-time student at a U.S. university should not infer that enrollment alone meets the citizenship or access requirements.
Keep a private readiness record listing evidence types and unresolved questions. You do not need to copy identity documents or private student records into a general job-search tracker to decide whether the program is worth investigating.
Match a research question to a current priority area
Use the 2026 Solicitation 2 priority descriptions, including their exclusions. The ASCR Computer Science area includes scientific computing and networking at extreme scales, with interests such as HPC storage and I/O, programming models, and resource management. “I write software” is too broad a research rationale.
The same area excludes several kinds of work, including discipline-specific analytics without a clear generalizable contribution to computing and certain computing research unconnected to supported scientific environments. A commercial dashboard or generic cloud project does not become a fit simply by adding an AI label.
Write a short chain: thesis question, proposed laboratory task, relevant priority aim, necessary laboratory capability, and evidence you expect to produce. Use that chain to test both the dissertation contribution and the need for the host before asking someone to support a proposal. It should not manufacture benchmark gains or claim access to a machine that nobody has secured.
The proposal guidelines require the project to be an integral part of the thesis, co-developed with the lab scientist in consultation with the primary advisor. They also require explicit use of host capabilities. Read the current full format instructions before preparing the final PDF; a polished statement of interest is not the same as that research proposal.
The laboratory scientist and thesis advisor have different roles
DOE's collaborator guidance assigns responsibilities to all three participants. The laboratory scientist helps develop the research, mentors the on-site work, and commits to securing necessary resources. The university advisor confirms the thesis connection and continues academic guidance.
A relevant publication identifies a possible collaborator; it does not establish their agreement, staff eligibility, or resource access for your proposed dates. Verify employment and availability throughout the proposed period. The guidance specifies qualifying laboratory research-staff arrangements and additional conditions for joint appointments and postdoctoral staff. Do not build nine months around someone whose committed laboratory tenure ends after six.
The university group does not have to hold Office of Science funding if the project otherwise meets the requirements. Existing support, however, needs a duplication check. A grant that already specifically funds the extended lab stay can create a different issue from a grant broadly supporting the thesis.
Both collaborators must approve the final proposal and provide their own support through the required system. A friendly conversation or a draft letter in your folder is not proof of receipt. Keep proposal version, approval status, and actual letter receipt as separate entries.
Check prior collaboration before assuming a return visit qualifies
The SCGSR FAQ and eligibility page restrict using the award to continue research at the same lab where the student is already conducting research. A candidate can investigate another participating host, subject to all other requirements. Do not treat strong existing relationships as an automatic eligibility advantage.
For past work, the eligibility page and applicant FAQ use “less than 3 months,” while some scientist-facing FAQ examples use “not more than 3 months.” If your history is exactly three months or involves multiple visits, ask the program how it counts the relevant experience rather than choosing the wording that produces your preferred answer.
Record dates, location, graduate-research scope, and collaborator accurately. Remote collaboration, co-located university space, and prior on-site research may require a fact-specific explanation. This article does not decide an individual boundary case or suggest omitting earlier work.
Worked asset one: a completed computing-thesis match
Mira is a fictional full-time computer-science PhD candidate with completed coursework and an approved thesis on failure-aware checkpoint placement in heterogeneous scientific workflows. Her invented history includes a two-month past graduate-research visit to the proposed host, no current on-site appointment, and no prior SCGSR award. Those assumptions support a screening exercise, not an eligibility ruling.
Her completed match record is below. The collaborator and resource arrangements are hypothetical; no real scientist was contacted and no allocation was reserved.
| Match field | Mira's completed research framing | Remaining verification |
|---|---|---|
| Thesis contribution | One dissertation chapter testing when checkpoint placement trades I/O overhead against recovery cost | Advisor approval of the bounded chapter scope |
| Priority connection | ASCR Computer Science: scientific HPC storage and I/O with a reusable scheduling method | Confirm exact current priority and exclusions |
| Laboratory reason | Compare behavior on heterogeneous scientific systems beyond her university's small test environment | Scientist confirms appropriate systems and access |
| Proposed task | Compare fixed-interval and byte-volume-triggered checkpoint policies; vary workload and failure assumptions; measure checkpoint time, lost work, and recovery cost | Agree datasets, experiment sizes, and resource limits |
| Evidence to produce | Reproducible experiment configurations, error analysis, and documented limitations | Confirm data/software review and dissemination procedures |
| Prior relationship | Two-month past visit recorded with dates and scope; not currently researching there | Program and host review actual history |
Her completed scope paragraph reads: “The laboratory period would investigate checkpoint-placement decisions under heterogeneous I/O conditions as part of my dissertation. I would compare fixed-interval and byte-volume-triggered baselines with a placement method that routes checkpoints toward less congested I/O paths. I would report recovery and overhead tradeoffs and identify workloads where the method fails. Laboratory expertise and agreed computing access would enable experiments unavailable in my current environment.”
This is a proposed experiment, not a claim that the method has improved performance. It names the comparison, intended evidence, thesis link, and unresolved access dependency. A request to gain supercomputing experience would still need those research and access details.
Mira rejects a different hypothetical project: rebuilding a group's general administrative dashboard. That task may be useful software work, but her record provides no demonstrated connection to the checkpoint thesis or the selected scientific-computing aim. She does not change the thesis description merely to match an available task.

Worked asset two: nine months with university milestones intact
For 2026 Solicitation 2, the current key-dates table gives a November 4, 2026 deadline at 5 p.m. Eastern and proposed starts from June 7 through October 4, 2027. SCGSR research periods run from three to twelve consecutive months. The start window is not the full duration of the award; research may continue for the approved consecutive period after starting.
Mira proposes July 6, 2027 through April 5, 2028 for this example. That is a nine-calendar-month planning interval with weekday endpoints, inside the published start window. Host holidays, resource availability, award terms, and approved dates still require confirmation. This is not an awarded period.
| Proposed phase | Research deliverable | University or program dependency |
|---|---|---|
| Months 1–2 | Validate configurations and establish the two baselines | Summer enrollment and agreed resource access |
| Months 3–5 | Implement the proposed method and conduct controlled comparisons | Fall enrollment; reserve time to prepare the required mid-term report |
| Months 6–8 | Sensitivity analysis and failure-case investigation | Spring enrollment; ongoing remote thesis-advisor review |
| Month 9 | Consolidate reproducibility materials and draft the thesis chapter | End-of-award report and planned return to the university |
Her home institution's fictional annual committee review falls during month five. She records a remote-review proposal requiring university approval, rather than a weekly commute back to campus. The lab's normal full-time research schedule must remain feasible.
Participant obligations require continued full-time graduate enrollment and full-time attendance at the host, with interruptions reported. Awards longer than six months require a mid-term report as well as the end report. Mira's November preparation task is an internal planning reminder; the program's assigned reporting instructions determine the actual due date.
Her financial worksheet separates support categories. The benefits page currently states a living stipend up to $3,600 monthly, assessed individually. Nine times that maximum is $32,400, a ceiling illustration rather than an award quote or money received.
Mira's invented living estimate is $3,250 monthly, or $29,250 across nine months. She separately lists $1,100 of inbound/outbound transportation. Current travel support is transportation-only, capped at $3,000 per award and contingent on the host being more than 50 miles from the home graduate institution. An estimated fare below the cap does not establish eligible transportation, approval, payment timing, or reimbursement already received.
Her separate, unfunded-cost ledger lists $3,600 in hypothetical university enrollment fees, $1,200 in research materials, $850 in conference travel, and $450 in moving services: $6,100 requiring separate arrangements. SCGSR does not add a general research-expense budget, and transportation reimbursement is not coverage for every relocation cost. Tuition and fees needed to remain enrolled remain the student and advisor's responsibility under program rules.
Mira's completed decision is “research plan bounded; keep enrollment support, concurrent funding, actual stipend, transportation eligibility, and excluded costs unresolved until confirmed.” She discloses other funding instead of assuming two sources can reimburse the same activity. The arithmetic is checked; funding approval is not invented.
Keep SCGSR separate from CSGF and submit a complete record
DOE CSGF is a different degree-support fellowship. Its benefits include tuition and required fees with support subject to renewal. Its 12-week laboratory practicum is outside thesis studies. SCGSR instead supports an integral thesis component for an already-qualified doctoral candidate. Do not import one program's eligibility, practicum purpose, or funding package into the other.
Before final submission, reconcile candidacy evidence, graduate transcripts, the final proposal, both collaborator approvals, and actual support-form receipts. The letters instructions require two online support forms, one from each responsible person, by the application deadline. A sent request is not a received form.
Use the completed research match and period plan to decide whether to proceed, narrow the project, or resolve a specific blocker. For a separate career search, compare recent technology openings on LandOffer. Neither a plausible thesis match nor an SCGSR award guarantees subsequent laboratory employment.
Sources and Further Reading
- DOE SCGSR: eligibility and candidacy proof
- DOE SCGSR: current research priorities
- DOE SCGSR: research proposal guidelines
- DOE SCGSR: laboratory scientist and thesis-advisor responsibilities
- DOE SCGSR: frequently asked questions
- DOE SCGSR: key dates
- DOE SCGSR: benefits
- DOE SCGSR: participant obligations
- DOE SCGSR: letters of support
- DOE CSGF: benefits
- DOE CSGF: laboratory practicum


