{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T11:06:04Z","timestamp":1790075164275,"version":"4.0.1"},"reference-count":48,"publisher":"Association for Computing Machinery (ACM)","issue":"PLDI","license":[{"start":{"date-parts":[[2026,6,8]],"date-time":"2026-06-08T00:00:00Z","timestamp":1780876800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["Proc. ACM Program. Lang."],"published-print":{"date-parts":[[2026,6,8]]},"abstract":"<jats:p>Data races, which occur when two or more threads incorrectly access the same memory location without appropriate synchronization, cause CUDA programmers considerable pain. Even experts struggle to reason about extreme parallelism, multiple memory spaces and scopes, and diverse synchronization mechanisms. Races can manifest as subtle data corruption, deadlock, or livelock, which are difficult to reproduce in a debugging environment. To date, there is no generally reliable tool for identifying races in GPU programs at scale. State-of-the-art dynamic GPU race detectors attempt to detect data races by faithfully tracking the memory operations, barriers, and memory fences executed by tens of thousands of threads. While this complexity enables the detection of subtle races, it comes with a cost: existing tools incur substantial memoryfootprint overheads (e.g., greater than 9\u00d7) and\/or runtime overheads (e.g., 1000\u00d7), and tend to report false positives. This paper introduces SuperCollider, a simple race detector for CUDA that tracks only threadlevel state, is agnostic to synchronization protocols, introduces no memory-footprint overhead, seamlessly handles shared, global, and tensor memory spaces, does not produce false positives, and is relatively fast. Although SuperCollider cannot detect all races, we demonstrate that it effectively catches many data races in professionally authored CUDA programs, including races that prior art cannot. Our approach exposed over 90 data races across libraries and CUDA benchmarks, and even found a race in the CUDA Programming Guide. Unlike prior art, SuperCollider can operate at the extreme scales required by modern GPU workloads.<\/jats:p>","DOI":"10.1145\/3808339","type":"journal-article","created":{"date-parts":[[2026,6,8]],"date-time":"2026-06-08T18:04:09Z","timestamp":1780941849000},"page":"2303-2327","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["SuperCollider: Scalable and Effective Data Race Detection for CUDA"],"prefix":"10.1145","volume":"10","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-1350-0165","authenticated-orcid":false,"given":"Mark","family":"Stephenson","sequence":"first","affiliation":[{"name":"NVIDIA, Austin, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6976-9000","authenticated-orcid":false,"given":"Sana","family":"Damani","sequence":"additional","affiliation":[{"name":"NVIDIA, Santa Clara, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6971-6996","authenticated-orcid":false,"given":"Mohamed","family":"Tarek Ibn Ziad","sequence":"additional","affiliation":[{"name":"NVIDIA, Westford, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-6712-0696","authenticated-orcid":false,"given":"Anis","family":"Ladram","sequence":"additional","affiliation":[{"name":"NVIDIA, New York, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6093-7602","authenticated-orcid":false,"given":"Michael","family":"Garland","sequence":"additional","affiliation":[{"name":"NVIDIA, Santa Clara, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2026,6,8]]},"reference":[{"key":"e_1_3_2_2_2","doi-asserted-by":"publisher","DOI":"10.1145\/2694344.2694391"},{"key":"e_1_3_2_3_2","doi-asserted-by":"publisher","unstructured":"Adam Betts Nathan Chong Alastair Donaldson Shaz Qadeer and Paul Thomson. 2012. GPUVerify: A Verifier for GPU Kernels. In Proceedings of the ACM International Conference on Object Oriented Programming Systems Languages and Applications. 113\u2013132. doi:10.1145\/2384616.2384625","DOI":"10.1145\/2384616.2384625"},{"key":"e_1_3_2_4_2","doi-asserted-by":"publisher","DOI":"10.1007\/s10703-023-00415-0"},{"key":"e_1_3_2_5_2","doi-asserted-by":"publisher","unstructured":"Ariel Eizenberg Yuanfeng Peng Toma Pigli William Mansky and Joseph Devietti. 2017. Barracuda: Binary-Level Analysis of Runtime Races in CUDA Programs. In Proceedings of the 38th ACM SIGPLAN Conference on Programming Language Design and Implementation. 126\u2013140. doi:10.1145\/3062341.3062342","DOI":"10.1145\/3062341.3062342"},{"key":"e_1_3_2_6_2","doi-asserted-by":"publisher","DOI":"10.5555\/1924943.1924954"},{"key":"e_1_3_2_7_2","doi-asserted-by":"publisher","DOI":"10.1145\/1806596.1806625"},{"key":"e_1_3_2_8_2","unstructured":"GCC. 2007. Bug 30475 - assert(int+100 > int) optimized away. https:\/\/gcc.gnu.org\/bugzilla\/show_bug.cgi?id=30475. Accessed: 10-30-2025."},{"key":"e_1_3_2_9_2","doi-asserted-by":"publisher","DOI":"10.1145\/3453953.3453972"},{"key":"e_1_3_2_10_2","unstructured":"John Hubbard Gonzalo Brito Chirayu Garg Nikolay Sakharnykh and Fred Oh. 2023. Simplifying GPU Application Development with Heterogeneous Memory Management. https:\/\/developer.nvidia.com\/blog\/simplifying-gpu-application-development-with-heterogeneous-memory-management\/."},{"key":"e_1_3_2_11_2","doi-asserted-by":"publisher","DOI":"10.1109\/SC41406.2024.00042"},{"key":"e_1_3_2_12_2","unstructured":"John Jacobson Martin Burtscher and Ganes Gopalakrishnan. 2024. HiRace-Artifact-SC24. https:\/\/github.com\/JohnJacobsonIII\/HiRace-Artifact-SC24\/."},{"key":"e_1_3_2_13_2","doi-asserted-by":"publisher","unstructured":"Zheming Jin and Jeffrey S. Vetter. 2023. A Benchmark Suite for Improving Performance Portability of the SYCL Programming Model. In 2023 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS). 325\u2013327. doi:10.1109\/ISPASS57527.2023.00041","DOI":"10.1109\/ISPASS57527.2023.00041"},{"key":"e_1_3_2_14_2","doi-asserted-by":"publisher","DOI":"10.1007\/11513988_49"},{"key":"e_1_3_2_15_2","doi-asserted-by":"publisher","DOI":"10.1145\/3477132.3483545"},{"key":"e_1_3_2_16_2","doi-asserted-by":"publisher","DOI":"10.1145\/2517349.2522736"},{"key":"e_1_3_2_17_2","doi-asserted-by":"publisher","DOI":"10.1145\/3062341.3062343"},{"key":"e_1_3_2_18_2","doi-asserted-by":"publisher","DOI":"10.1109\/MC.1993.274940"},{"key":"e_1_3_2_19_2","doi-asserted-by":"publisher","DOI":"10.1145\/3575693.3575750"},{"key":"e_1_3_2_20_2","doi-asserted-by":"publisher","unstructured":"Guodong Li Peng Li Geof Sawaya Ganesh Gopalakrishnan Indradeep Ghosh and Sreeranga P Rajan. 2012. GKLEE. In Proceedings of the 17th ACM SIGPLAN symposium on Principles and Practice of Parallel Programming. ACM. doi:10.1145\/2145816.2145844","DOI":"10.1145\/2145816.2145844"},{"key":"e_1_3_2_21_2","doi-asserted-by":"publisher","DOI":"10.1145\/781131.781148"},{"key":"e_1_3_2_22_2","doi-asserted-by":"publisher","DOI":"10.1145\/3665251"},{"key":"e_1_3_2_23_2","doi-asserted-by":"publisher","DOI":"10.1145\/3297858.3304043"},{"key":"e_1_3_2_24_2","unstructured":"NVIDIA. 2025. Compute Sanitizer v2025.3.1. https:\/\/docs.nvidia.com\/cuda\/sanitizer-docs\/pdf\/ComputeSanitizer.pdf"},{"key":"e_1_3_2_25_2","unstructured":"NVIDIA. 2025. CUDA C++ Programming Guide. https:\/\/docs.nvidia.com\/cuda\/cuda-c-programming-guide\/. Accessed: 9-5-2025."},{"key":"e_1_3_2_26_2","unstructured":"NVIDIA. 2025. CUDA C++ Programming Guide : Atomic Functions. https:\/\/docs.nvidia.com\/cuda\/cuda-c-programming-guide\/#atomic-functions. Accessed: 11-13-2025."},{"key":"e_1_3_2_27_2","unstructured":"NVIDIA. 2025. CUDA C++ Programming Guide : Distributed Shared Memory. https:\/\/docs.nvidia.com\/cuda\/cuda-c-programming-guide\/index.html#distributed-shared-memory. Accessed: 11-12-2025."},{"key":"e_1_3_2_28_2","unstructured":"NVIDIA. 2025. CUTLASS 4.3.0. https:\/\/github.com\/NVIDIA\/cutlass"},{"key":"e_1_3_2_29_2","unstructured":"NVIDIA. 2025. NVIDIA CUDA Profiling Tools Interface (CUPTI) - CUDA Toolkit. https:\/\/developer.nvidia.com\/cupti"},{"key":"e_1_3_2_30_2","unstructured":"NVIDIA. 2025. NVIDIA\/cuda-samples. https:\/\/github.com\/NVIDIA\/cuda-samples"},{"key":"e_1_3_2_31_2","unstructured":"NVIDIA. 2025. PTX ISA 9.0 Documentation - Instruction Operands : Generic Addressing. https:\/\/docs.nvidia.com\/cuda\/parallel-thread-execution\/#generic-addressing"},{"key":"e_1_3_2_32_2","unstructured":"NVIDIA. 2025. PTX ISA 9.0 Documentation - Memory Consistency Model. https:\/\/docs.nvidia.com\/cuda\/parallel-thread-execution\/#morally-strong-operations"},{"key":"e_1_3_2_33_2","unstructured":"NVIDIA. 2025. PTX ISA 9.0 Documentation - Miscellaneous Instructions : nanosleep. https:\/\/docs.nvidia.com\/cuda\/parallel-thread-execution\/#miscellaneous-instructions-nanosleep"},{"key":"e_1_3_2_34_2","unstructured":"NVIDIA. 2025. PTX ISA 9.0 Documentation - Parallel Synchronization and Communication Instructions. https:\/\/docs.nvidia.com\/cuda\/parallel-thread-execution\/#parallel-synchronization-and-communication-instructions-match-sync"},{"key":"e_1_3_2_35_2","unstructured":"NVIDIA Corporation. 2017. NVIDIA Tesla V100 GPU Architecture. https:\/\/images.nvidia.com\/content\/volta-architecture\/pdf\/volta-architecture-whitepaper.pdf.Whitepaper."},{"key":"e_1_3_2_36_2","unstructured":"NVIDIA Corporation. 2022. NVIDIA H100 Tensor Core GPU Architecture Whitepaper. https:\/\/resources.nvidia.com\/en-us-hopper-architecture\/nvidia-h100-tensor-c.Whitepaper."},{"key":"e_1_3_2_37_2","doi-asserted-by":"publisher","DOI":"10.1145\/3366428.3380768"},{"key":"e_1_3_2_38_2","doi-asserted-by":"publisher","DOI":"10.1145\/3296979.3192368"},{"key":"e_1_3_2_39_2","unstructured":"John Regehr. 2010. A Guide to Undefined Behavior in C and C++. https:\/\/blog.regehr.org\/archives\/213. Accessed: 10-30-2025."},{"key":"e_1_3_2_40_2","doi-asserted-by":"publisher","DOI":"10.1145\/265924.265927"},{"key":"e_1_3_2_41_2","volume-title":"Proceedings of the 2012 USENIX Conference on Annual Technical Conference","author":"Serebryany Konstantin","year":"2012","unstructured":"Konstantin Serebryany, Derek Bruening, Alexander Potapenko, and Dmitry Vyukov. 2012. AddressSanitizer: A Fast Address Sanity Checker. In Proceedings of the 2012 USENIX Conference on Annual Technical Conference (Boston, MA) (USENIX ATC\u201912). USENIX Association, USA, 28. https:\/\/www.usenix.org\/conference\/atc12\/addresssanitizer-fast-address-sanity-checker"},{"key":"e_1_3_2_42_2","doi-asserted-by":"publisher","unstructured":"Konstantin Serebryany and Timur Iskhodzhanov. 2009. ThreadSanitizer: Data Race Detection in Practice. In Proceedings of the Workshop on Binary Instrumentation and Applications. 62\u201371. doi:10.1145\/1791194.1791203","DOI":"10.1145\/1791194.1791203"},{"key":"e_1_3_2_43_2","doi-asserted-by":"publisher","DOI":"10.1145\/3639477.3640328"},{"key":"e_1_3_2_44_2","doi-asserted-by":"publisher","DOI":"10.1145\/2908080.2908114"},{"key":"e_1_3_2_45_2","doi-asserted-by":"publisher","DOI":"10.5281\/zenodo.19058944"},{"key":"e_1_3_2_46_2","unstructured":"Texas State University. 2019. ECL graphs. https:\/\/userweb.cs.txstate.edu\/~burtscher\/research\/ECLgraph\/index.html."},{"key":"e_1_3_2_47_2","doi-asserted-by":"publisher","unstructured":"Oreste Villa Mark Stephenson David Nellans and Stephen W Keckler. 2019. NVBit: A Dynamic Binary Instrumentation Framework for NVIDIA GPUs. In Proceedings of the 52nd Annual IEEE\/ACM International Symposium on Microarchitecture. ACM. doi:10.1145\/3352460.3358307","DOI":"10.1145\/3352460.3358307"},{"key":"e_1_3_2_48_2","doi-asserted-by":"publisher","DOI":"10.1145\/2349896.2349905"},{"key":"e_1_3_2_49_2","doi-asserted-by":"publisher","DOI":"10.1145\/3016078.2851145"}],"container-title":["Proceedings of the ACM on Programming Languages"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3808339","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T10:39:36Z","timestamp":1790073576000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3808339"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,8]]},"references-count":48,"journal-issue":{"issue":"PLDI","published-print":{"date-parts":[[2026,6,8]]}},"alternative-id":["10.1145\/3808339"],"URL":"https:\/\/doi.org\/10.1145\/3808339","relation":{},"ISSN":["2475-1421"],"issn-type":[{"value":"2475-1421","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,8]]},"assertion":[{"value":"2025-11-13","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-04-03","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-06-08","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}