Papers › Comet: Fine-grained Computation-communication Overlapping for Mixture-of-Experts

Comet: Fine-grained Computation-communication Overlapping for Mixture-of-Experts

27 Feb 2025arXiv:2502.19811archive 2025-07-28

Shulai Zhang, Ningxin Zheng, Haibin Lin, Ziheng Jiang, Wenlei Bao, Chengquan Jiang, Qi Hou, Weihao Cui, Size Zheng, Li-Wen Chang, Quan Chen, Xin Liu

Mixture-of-experts (MoE) has been extensively employed to scale large language models to trillion-plus parameters while maintaining a fixed computational cost. The development of large MoE models in the distributed scenario encounters the problem of large communication overhead. The inter-device communication of a MoE layer can occupy 47% time of the entire model execution with popular models and frameworks. Therefore, existing methods suggest the communication in a MoE layer to be pipelined with the computation for overlapping. However, these coarse grained overlapping schemes introduce a notable impairment of computational efficiency and the latency concealing is sub-optimal. To this end, we present COMET, an optimized MoE system with fine-grained communication-computation overlapping. Leveraging data dependency analysis and task rescheduling, COMET achieves precise fine-grained overlapping of communication and computation. Through adaptive workload assignment, COMET effectively eliminates fine-grained communication bottlenecks and enhances its adaptability across various scenarios. Our evaluation shows that COMET accelerates the execution of a single MoE layer by 1.96× and for end-to-end execution, COMET delivers a 1.71× speedup on average. COMET has been adopted in the production environment of clusters with ten-thousand-scale of GPUs, achieving savings of millions of GPU hours.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 20 code samples harvested from 2 repositories linked to this paper: 9 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 0 violated, 8 with no contract checked; 1 where Syntology's instrument failed) · 11 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 17 samples from 1 repository, 6 ran (6 with no instrument failure; 0 where Syntology's instrument failed); community (archive-listed): 3 samples from 1 repository, 3 ran (2 with no instrument failure; 1 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

No sample is linked to this paper from 1 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-24 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

bytedance/flux officialmentioned in papermentioned on GitHubpytorchApache-2.0Syntology: 17 samples, 6 ran (0 instrument-failed) · 11 unverified report
multi-swe-bench/magentless mentioned on GitHubMITSyntology: 3 samples, 3 ran (1 instrument-failed) · 0 unverified report
multi-swe-bench/mopenhands mentioned on GitHubMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Multi-SWE-bench: A Multilingual Benchmark for Issue Resolving (arXiv:2504.02605; that paper's own run record: 4 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 1 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
multi-swe-bench/mswe-agent mentioned on GitHubSyntology: not harvested report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

20 samples harvested: 9 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 0 violated, 8 with no contract checked; 1 where Syntology's instrument failed) · 11 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · our draft was wrong
8ran
11unverified

Licence: 0 of the 20 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 2 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

bytedance/flux official repository 17 samples, 6 ran (0 instrument-failed) · 11 unverified
calc_scatter_index_stable python/flux/testing/moe_utils.py ran fingerprinted Apache-2.0 (permissive) · c252afed940d6e30 · report fixture: signature choosed_experts:Tensor/2/int · returned Tensor · level recorded 2026-09-24
get_inter_node_ag_group python/flux/ag_kernel_internode.py ran Apache-2.0 (permissive) · 3e48284521d53051 · report fixture: signature tp_group:MockProcessGroup/0/?, nnodes:int/0/int · returned list · level recorded 2026-09-24
get_torch_prof_ctx python/flux/util.py ran Apache-2.0 (permissive) · 318a824a192ac02f · report fixture: signature do_prof:bool/0/bool · returned nullcontext · level recorded 2026-09-24
is_fp8_dtype python/flux/testing/gpu_perf_model.py ran Apache-2.0 (permissive) · f47e5cdd7171b461 · report fixture: signature dtype:dtype/0/? · returned bool · level recorded 2026-09-24
is_fp8_dtype python/flux/util.py ran Apache-2.0 (permissive) · 6469bfc20e491271 · report fixture: signature dtype:dtype/0/? · returned bool · level recorded 2026-09-24
torch_pre_attn_all_to_all_transpose python/flux/testing/ulysses_sp_utils.py ran Apache-2.0 (permissive) · 36283518ba9f282d · report fixture: signature sp_group:MockProcessGroup/0/?, input:Tensor/1/float, bs:int/0/int, seq_len:int/0/int, nh:int/0/int, head_dim:int/0/int, seq_lens_cpu:NoneType/0/? · returned list · level recorded 2026-09-24
bench_func python/flux/util.py unverified Apache-2.0 (permissive) · b2e2521687ac95ce · report
calc_gather_index python/flux/testing/moe_utils.py unverified Apache-2.0 (permissive) · b0bd0e05915bf19c · report
calc_gather_index_stable python/flux/testing/moe_utils.py unverified Apache-2.0 (permissive) · c7158b2a6c069b16 · report
get_device_property python/flux/testing/gpu_perf_model.py unverified Apache-2.0 (permissive) · bdad45577321a1d2 · report
get_dist_env python/flux/dist_utils.py unverified Apache-2.0 (permissive) · c23cadd88b362499 · report
get_intra_node_ag_group python/flux/ag_kernel_internode.py unverified Apache-2.0 (permissive) · 80683aed2fa79e07 · report
get_intra_node_pg_group python/flux/gemm_rs_sm80.py unverified Apache-2.0 (permissive) · 673d58d820da47c9 · report
nvsmi python/flux/testing/gpu_perf_model.py unverified Apache-2.0 (permissive) · 0c0fc4a132ede3cb · report
run_perf python/flux/testing/utils.py unverified Apache-2.0 (permissive) · 146ee989a1ad16cb · report
torch_post_attn_all_to_all_transpose python/flux/testing/ulysses_sp_utils.py unverified Apache-2.0 (permissive) · 24ef3f53f604fc80 · report
torch_pre_attn_qkv_pack_a2a python/flux/testing/ulysses_sp_utils.py unverified Apache-2.0 (permissive) · e693c5491138e1e3 · report
multi-swe-bench/magentless community (archive-listed) 3 samples, 3 ran (1 instrument-failed) · 0 unverified
construct_file_meta_data agentless/fl/Index.py ran MIT (permissive) · 4dc6308a692bee4f · report fixture: signature file_name:str/0/str, clazzes:list/1/dict, functions:list/1/dict · returned dict · level recorded 2026-09-24
parse_patch get_repo_structure/get_patch_info.py ran · our draft was wrong fingerprinted MIT (permissive) · f2c4e9391d254c7e · report fixture: signature patch:str/0/str · returned list · level recorded 2026-09-25
parse_python_file get_repo_structure/get_repo_structure.py ran fingerprinted MIT (permissive) · 94dc3cbe589f0afa · report fixture: signature file_path:str/0/str, file_content:str/0/str · returned tuple · level recorded 2026-09-25

Tasks

Computational EfficiencyMixture-of-Experts

1 archive task tag without a task page not shown.

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

MoE

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections