ATC 2026

• ATC 2026 is now an ACM conference, but with the same community and scope as before.
• Submit a two-page extended abstract for an initial round of reviews.
• There is no abstract registration period. Submit abstracts, extended abstracts, and full papers together by the submission deadline.
• ATC papers will be published under ACM's Open Access program.

The scope of ATC covers all practical aspects related to computer systems, including but not limited to: operating systems; runtime systems; parallel and distributed systems; storage; networking; ML for system and systems for ML; security and privacy; virtualization; software-hardware interactions; performance evaluation and workload characterization; reliability, availability, and scalability; energy and power management; and bug-finding, tracing, analyzing, and troubleshooting. We value submissions more highly if they are accompanied by clearly defined artifacts not previously available, including traces, original data, source code, or tools developed as part of the submitted work. We particularly encourage new ideas and approaches.

Submissions will be judged on novelty, significance, interest, clarity, relevance, and correctness. A good paper will:

• Motivate a significant problem;
• Propose and implement an interesting, compelling solution;
• Demonstrate the practicality and benefits of the solution;
• Draw appropriate conclusions;
• Clearly describe the paper's contributions;
• Articulate the advances beyond previous work; and
• Be accessible to the broader systems community.

We encourage submission of groundbreaking work in significant new directions, with the understanding that the evaluation criteria for papers addressing new problems may be different from those continuing a line of work in a more established area.

Submissions should contain original, unpublished material. Simultaneous submission of the same work to multiple venues, submission of previously published work, or plagiarism is not allowed.

Submissions that extend the authors' previous work are welcome, but authors must explain the differences between the ATC submission and the prior work, much in the same way that the authors are expected to articulate the contributions of their submission as they relate to prior work by others.

Prior or concurrent workshop publication does not preclude publishing a related paper in ATC. As long as there is significant additional content in the submission compared to the prior workshop publication, the PC will evaluate the submission's entire contribution, not just the delta.

Prior or concurrent publication in non-peer-reviewed contexts, like arXiv.org, technical reports, talks, and social media posts, is explicitly permitted.

Submitting a paper that had been previously submitted to and not accepted by another conference is permitted, although authors are expected to have improved the paper to address substantive issues raised in previous reviews. Authors should provide information regarding the previous submission(s) and a summary of the subsequent revisions to the paper. This description, which will be supplied to reviewers after they’ve completed their reviews, helps reviewers who may have reviewed a previous draft of the work to appreciate any improvements to the currently submitted work. Please try to limit the description of changes to one page. All information should be properly anonymized, as described below, and should be uploaded via the submission form.

ATC uses double-blind reviewing. Authors should make a good faith attempt to anonymize your submission. Avoid identifying yourself or your institution explicitly or by implication (e.g., through the references or acknowledgments). The first page should use the paper ID assigned during registration in place of the author names. If the name of your project or system is already known to the community (e.g., through arXiv, technical reports, talks, social media posts, or other uses), your submission must use an anonymized name.

Use care in referring to your own related work. Do not omit references to your prior work, as this would make it difficult for reviewers to place your submission in its proper context. Instead, reference your past work in the third person, just as you would any other piece of related work. For example, you might say "Our system modifies the XYZ operating system built by Lee et al. [Lee17]".

For concurrent submissions on related topics, cite an anonymized version of the concurrent submission and discuss the relation between the submissions. Additionally, email the PC chairs with a copy of your other concurrent submission.

For submissions to the operational systems track, which report on experiences with a system at your institution or organization, you should refer to the system anonymously but describe the properties of the system that are needed to appreciate the work (e.g., size of the user base, volume of requests, etc.). We recognize that, in some cases, these properties may allow a reviewer to identify your institution or organization.

In addition to the paper, authors must submit a two-page extended abstract that provides an overview of the motivation and the contribution of the paper. The extended abstract will be reviewed against the following criteria:

• Be within the scope of the conference;
• Address an interesting and important problem;
• Advance the state-of-the-art; and
• Convince that the experimental results properly support the ideas of the paper.

The extended abstract has the same formatting requirements as the main paper (see below). It should include the title block, but it need not have an abstract. Everything except the references count towards the two-page limit. If the extended abstract violates the formatting requirements or exceeds the page limit, the submission will be rejected.

Authors can use a copy of the introduction as the extended abstract (if it fits in 2 pages). However, both the extended abstract and full submission must stand by themselves without each other (e.g., the full submission itself should have an introduction), as reviewers may read only one of them depending on the review round. See additional guidelines for authors here.

The extended abstract must be submitted together with the paper in HotCRP by the submission deadline. The extended abstract is intended for review purposes and will not be published.

ATC accepts both long and short submissions. Long submissions must not exceed 12 pages, and short submissions must not exceed 6 pages. In both cases, the page limit excludes references and appendices but includes all text, figures, tables, footnotes, etc. For both types, authors should submit an extended abstract. Both types are reviewed to the same standards and differ primarily in scope. A short submission presents a complete idea that is properly evaluated, as a long submission. Optionally, authors can include an appendix to provide material that may be of interest to some reviewers, such as algorithm details, proofs, and code snippets, but is not central to the contributions of the paper. For further details, see the submission instructions.

ATC solicits papers that describe the design, implementation, analysis, and experience with real-world deployment of systems and networks. Papers for the operational systems track need not present new ideas or results to be accepted but should convey practical insights. Note that the rules regarding submission and anonymization are different for this track. The final program will explicitly identify papers accepted to the operational systems track to distinguish them from papers accepted to the regular track.

ATC will provide an opportunity for authors to respond to reviews prior to final consideration of the papers at the program committee meeting. Authors must limit their responses to (a) correcting factual errors in the reviews or (b) directly addressing questions posed by reviewers. In particular, responses must not include new experiments or data, describe additional work completed since submission, or promise additional work to follow. As PC members are not required to review supplementary materials, responses must not rely on the existence of those materials.

We have made some modifications to the review process to address the increasing number of submissions and reviewing load in our community. While there is no perfect review process, we believe the new process is an improvement. Generally speaking, the goal of the review process is to (1) select the best papers for publication and (2) provide meaningful feedback to authors. The process used in prior years, where each paper receives at least three full reviews, generated excessive review load, required increasingly large PCs, and resulted in inconsistent review quality. The new process aims to lessen review load, reduce PC size, and improve the consistency of reviews.

Submissions will undergo two rounds of reviews. The first round will evaluate if the extended abstract meets the criteria above (see “Extended Abstract”). To do so, each extended abstract will be read by two experienced reviewers with a broad perspective on systems research. Reviewers will answer the following questions in their reviews. Papers that fail to meet the required criteria will be rejected early without additional reviewing. We will strive to provide meaningful feedback to such papers. The other papers will advance to the second round, where 3-4 reviewers will evaluate the full paper. Then, reviews will be made available to authors. For submissions that were not rejected early, authors will be able to respond (see “Author Response Period” above). Subsequently, reviewers will reach a decision based on an online discussion.

Papers will be submitted electronically in PDF format.

Long submissions may have at most 12 pages, while short submissions may have at most 6 pages, including all text, figures, tables, etc. Bibliographic references are not included in the 6- or 12-page limit. Use A4 or US letter paper size, with all text and figures fitting inside a 178 x 229 mm (7 x 9 in) block centered on the page, using two columns separated by 8 mm (0.33 in) of whitespace. Use 10-point font (typeface Times Roman or similar) on 12-point (single-spaced) leading. Graphs and figures should be readable without magnification; they are encouraged to be in color, but should remain readable if printed in grayscale. All pages should be numbered, and references within the paper should be hyperlinked.

Submissions violating these rules will not be considered for publication, and there will be no extensions for fixing violations. We encourage you to upload an early draft of the paper well before the deadline to check if the paper meets the formatting rules.

Most of these rules are automatically applied when using the official SIGPLAN LaTeX or MS Word templates from the ACM.

For Latex, we recommend you use:

\documentclass[sigplan,10pt]{acmart}
\renewcommand\footnotetextcopyrightpermission[1]{}
\pagestyle{plain}
...
\settopmatter{printfolios=true}
\maketitle

Reviewing will be done mostly by members of the program committee, with limited use of outside reviewers. Submissions will be treated as confidential; however, papers accompanied by nondisclosure agreement forms will not be considered for publication.

Authors must not submit papers produced entirely or substantively by generative AI. Authors must provide a statement attesting to that, as part of the submission process in HotCRP. For any questions about the appropriate use of generative AI, contact the chairs.

To avoid conflicts of interest in the review process, when you register and submit your paper, we ask that you provide information about conflicts between any of the authors of your submission and PC members. Use the following guidelines to determine conflicts:

Institutional: You are currently employed at the same institution, have been previously employed at the same institution within the past two years, or are going to begin employment at the same institution. Former interns should list their internship supervisor(s) and collaborator(s) as conflicts, not the entire institution where they interned.

Advisor: You have a past or present association as PhD thesis advisor or advisee.

Collaborator: You have a collaboration on a project, publication, grant proposal, or editorship within the past two years.

Personal: You are close family relatives (e.g., spouses, domestic partners, parents, children, or siblings).

The PC chairs will review paper conflicts to ensure the integrity of the reviewing process, adding conflicts if necessary. You should not improperly identify a PC member as a conflict in an attempt to avoid having the individual review your paper.

PC members will not be able to review, read the reviews of, or participate in discussions of papers they are conflicted with. The review process for papers conflicted with both PC chairs will be managed by another PC member designated as the "conflict chair". PC chairs will not submit papers.

Papers selected by the program committee will receive conditional acceptance, with the final acceptance subject to revisions and approval by a program committee member acting as a shepherd. Authors of accepted papers will be expected to supply electronic versions of their papers and encouraged to supply source code and raw data to help others replicate and understand their results, as part of an artifact evaluation process. To facilitate broad technical discussion, all accepted papers will be made available online in advance of the conference. The official publication date will be the date the proceedings are made publicly accessible. For more information about the ACM OpenSurround Service, please go to https://www.acm.org/publications/policies/free-access. Papers of particular merit will be forwarded to the ACM Transactions on Computer Systems (TOCS) and the Communications of the ACM's Research Highlights (CACM RH) for possible publication.

Starting January 1, 2026, ACM will transition to Open Access. That means ATC publications will be Open Access. Authors have two options for publication: the ACM Open institutional model or by paying Article Processing Charges (APCs). With over 2,600 institutions already part of ACM Open, the majority of ACM-sponsored conference papers will not require APCs. Authors from institutions not in ACM Open will need to pay an APC to publish, unless they qualify for a waiver. Waivers are rare and are based on criteria set by ACM. As this change could present financial challenges, ACM has approved a subsidy for 2026 to ease the transition and allow time for institutions to join ACM Open. The subsidy will offer $250 APC for ACM/SIG members and $350 APC for non-members. This represents a 65% discount.

Papers must be submitted electronically in PDF format via the web submission form.

Any questions about paper submissions can be emailed to the PC chairs.