ISCA 2027

The International Symposium on Computer Architecture (ISCA) is the premier forum for new ideas and experimental results in computer architecture. The conference specifically seeks particularly forward-looking and novel submissions. In 2027, the 54^th edition of ISCA will be held in Atlanta at the Atlanta Hyatt Regency from June 5 to June 9, 2027. ISCA 2027 will be co-located with the Federated Computing Research Conference (FCRC).

As with previous editions of ISCA, there will be a main track and an industry track. For the main track, the dates, topics, and review policies are presented here. Industry-track details and dates will be announced separately.

ISCA-54 will be conducting Artifact Evaluation. Authors of accepted papers are encouraged to submit their artifacts for evaluation.

Topics

Papers are solicited on a broad range of topics, including (but not limited to):

• Processor cores and microarchitecture
• Parallel and multicore systems
• Memory systems, storage, caches, DRAM, HBM, and emerging memories
• Accelerators, GPUs, TPUs, reconfigurable and domain-specific architectures
• Machine learning for architecture and architectures for machine learning
• Datacenter, cloud-scale, serverless, and disaggregated systems
• IoT, mobile, edge, and embedded architectures
• Interconnection networks, chiplets, and die-to-die fabrics
• Security, privacy, virtual memory, and virtualization
• Power, energy, thermal management, and sustainable computing
• Reliability, availability, verification, testing, and correctness
• Quantum, analog, in-memory, and other emerging computing
• Programming languages, software development, and accelerator interfaces
• Modeling, simulation, benchmarking, methodologies, tools, and measurement
• Architectures for emerging applications, including machine learning and bioinformatics
• Human factors and user studies

Submission and Review Process

Submissions undergo a strict Phase 0 compliance check covering formatting (including figure and table font size and readability), page limits, anonymization, concurrent submissions, prior publication, and generative-AI disclosure. Clear and verifiable violations are desk-rejected; formatting and figure-font rule breaches are enforced, not left to reviewer discretion.

Each compliant paper initially receives three reviews. First-round decisions are never made on fewer than three reviews. Papers advancing from the first round receive one additional review (four reviews in total) and enter a rebuttal/revision period. Detailed page-limit, anonymity, conflict-of-interest, and submission-system instructions will be published separately.

Limit on Submissions per Author

Each author may be listed on at most 8 submissions to the ISCA 2027 main track. Submissions will be screened after the abstract deadline. If an author exceeds the limit, the Program Chairs will desk-reject that author’s excess submissions until the limit is met, selecting them at random. Authors added after abstract registration also count toward this limit.

Resubmissions

Authors of a submission that was previously reviewed and rejected at ISCA, HPCA, ASPLOS, MICRO, or any other venue must upload a separate note at submission time describing the changes since the previous submission(s). This note helps reviewers who may have seen an earlier draft understand how the work has improved. Please try to keep this note to one page. The note is not counted toward the page limit and is visible only to reviewers. Resubmissions that show no substantial improvement over the previously reviewed version, or that omit the required change note, are subject to desk rejection.

New Policies Regarding Expected Reviewer Service for Authors

To keep high review quality and ensure a productive and pleasant reviewer experience in anticipation of increasing paper submissions, for ISCA-54, we are continuing expectations for authors regarding program-committee service and reserve reviewing.

PC Service for Authors Submitting More than 6 Abstracts

Any senior author (i.e., holding a PhD or having equivalent level of seniority) registering more than 6 abstracts (regardless of whether any of those abstracts are later withdrawn) may be called upon to serve on the PC and cannot decline. Another senior author may be nominated in place. Final decisions on this matter will be at the discretion of the PC chairs.

Reserve Reviewers

To prepare for the possibility of a higher volume of submissions, we require at least one senior author (as defined above) per paper must register as a reserve reviewer (unless exempt under the criteria below). [We thank the OOPSLA program chairs for piloting this practice!]

The goal of this policy is to uphold the high standard of reviews within the SIGARCH/TCCA community. To achieve this, we must ensure manageable review loads, prevent burnout, and encourage reviewers to stay engaged for future rounds. High-quality reviews are one of the community’s greatest assets, playing a crucial role in elevating the quality of research for everyone.

Our hope is that these reserve reviewers won’t be needed at all! Those selected to serve as a reviewer will be acknowledged in the proceedings.

When registering a paper in HotCRP, the form will include a field for the designated reserve reviewer. After the final paper submission, we will notify reserve reviewers if their reviewing services will be needed, and they will be expected to enter their areas of expertise in HotCRP in a timely fashion and update their COIs.

We define “senior” authors as those who completed their PhD, or equivalent industry experience. A paper is exempt from the reserve reviewer policy if:

• The paper has no senior authors.
• At least one senior author is already on the program committee.
• At least one senior author is subject to the 6-abstract criterion above.
• Every senior author of the paper satisfies one or more of these criteria:
  • Has never published in a conference sponsored by SIGARCH or TCCA.
  • Is (co)chairing a SIGARCH or TCCA conference with 150 or more submissions last year, this year, or next year.
  • Has some other exceptional circumstance that didn’t prevent writing the paper but prevents doing any reviewing. This must be cleared at least three days before submission with the PC Chairs.

It is okay for the same person to serve as the reserve reviewer for more than one paper. Please enter their information for each such paper (preferably identically). For cases in which a paper is exempt according to the above criteria, please enter “exempt” in the reserve reviewer field.

Abstract Registration and Paper Finalization

The paper title, abstract, and full author list must be final at abstract registration (November 9, 2026), one week before the full paper deadline. Authorship cannot be changed after abstract registration, except to correct an honest error with the approval of the Program Chairs. Placeholder or “ghost” abstracts are not allowed. Submissions whose final paper deviates substantially from the registered title or abstract may be desk-rejected.

Generative AI and LLM Use Policy

ISCA 2027 follows the ACM/IEEE generative AI baseline for publications. Language editing, code assistance, and literature search do not require disclosure. Any use that contributes substantive technical content must be disclosed in an acknowledgements-style statement (for example, generated scientific prose, related-work discussion, or experimental data). Authors remain responsible for correctness, originality, attribution, and integrity. Undisclosed substantive generation is a policy violation. Clear, verifiable issues (e.g., fabricated references) may lead to desk rejection or misconduct handling under ACM/IEEE procedures.

LLM use in the review process is consistent with ACM’s peer-review policies. In particular, papers must not be uploaded to a third-party LLM that does not maintain confidentiality.

Concurrent Submissions & Workshops Policy

By submitting a manuscript to ISCA 2027, the authors guarantee that the manuscript has not been previously published or accepted for publication in a substantially similar form in any conference, journal, or the archived proceedings of a workshop (e.g., in the ACM/IEEE digital library) — see exceptions below. The authors also guarantee that no paper that contains significant overlap with the contributions of the submitted paper will be under review for any other conference or journal or an archived proceedings of a workshop during the ISCA 2027 review period. Violation of any of these conditions will lead to rejection.

The only exceptions to the above rules are for the authors' own papers in:

• workshops without archived proceedings such as in the ACM/IEEE digital library (or where the authors chose not to have their paper appear in the archived proceedings), or
• venues such as IEEE CAL or arXiv where there is an explicit policy that such publication does not preclude longer conference submissions.

In all such cases, manuscripts may still be submitted, but should:

• Use a different title for the version you submit to ISCA, and
• Provide this information on the submission form – the PC chair will make this information available to reviewers if it becomes necessary to ensure a fair review.

In addition, it is important to note that ACM/IEEE prohibit authors from reusing their own text or figures without attribution; doing otherwise either violates rules regarding plagiarism or anonymity. This applies to published works only, i.e. not IEEE CAL or arXiv works.

As always, if you are in doubt, it is best to contact the program chairs.

Plagiarism Policies

The following cover a range of ethical issues concerning the misrepresentation of other works or one's own work.

• ACM Plagiarism Policy
• IEEE Plagiarism Policy

Code of Conduct

All participants are expected to follow applicable ACM and IEEE professional conduct and nondiscrimination policies.

Organizers

General Co-Chairs	Hyesoon Kim Georgia Tech
	Prashant Nair University of British Columbia
Program Chair	Yuan Xie HKUST
Vice Program Chair	Jishen Zhao University of California, San Diego
Industry Track Chair	Pradip Bose IBM