[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82673-en":3,"doc-seo-82673-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},82673,3848291630094,"Emma Wilson","https://eur-avatar.wpscdn.com/davatar_085a072bc5b1113ac321206ff7593b45",8,"Research & Report","Static PTX Metrics Track Structural Kernel Regressions but Miss Semantic Ones","The work pairs static PTX metrics for each GPU kernel—register count, spill count, and instruction count—with CUDA-event-timed runtime measured across five GPU classes (RTX 3060, A10, L40S, A100 SXM4, H100 NVL). Results show that, within the provided corpus/toolchain, static and measured signals separate along one axis: structural bug variants create nonzero static deltas, while semantic constant-only changes yield zero static delta. Runtime deltas at the sub-millisecond scale are dominated by noise and hardware/variance sensitivity.","arXiv :2607 .02541v1 [ cs .DC] 23 Jun 2026  \nStatic PTX Metrics Track Structural Kernel Regressions but Miss Semantic Ones  \nDipankar Sarkar[0000-0001-5431-6367]  \nArizona State University, USA  \n[dsarkar3@asu.edu](dsarkar3@asu.edu)  \nAbstract. We pair each GPU kernel’s static PTX metrics (registers, spills, instruction count) with CUDA-event-timed runtime on five GPU classes: RTX 3060, A10, L40S, A100 SXM4, and H100 NVL. In this corpus and toolchain the static and measured signals separate cleanly along one axis. Per-pair ∆regs and ∆instrs are identical across all five GPUs for any given (correct, buggy) pair. Measured ∆perf% is not.  \nStructural bugs that change the kernel’s work are unambiguous in the static signal. The gelu triton buggy variant, which drops a leading 0 .5 factor, removes 8 instructions and 8 registers. The corresponding measured ∆perf% on RTX 3060 is +3 .2%, within the run-to-run noise band atthe sub-millisecond scale these corpus kernels occupy. Semantic bugs that swap one constant for another are invisible to the static signal. The softmax triton buggy variant, which substitutes other=0 .0 for-inf on the masked load, compiles to byte-identical PTX. The paper’s bounded claim is that, for this corpus and toolchain, a static-PTX delta gate is a portable pre-filter that separates structural from semantic changes;  \nmeasured runtime deltas at this scale are hardware-and noise-sensitive and are not a substitute.  \nKeywords: GPU compilation · static analysis · PTX · performance regression · CI gating  \n1 Introduction  \nCI gating on GPU kernel changes is expensive. Running every variant on real hardware costs seconds to minutes per test. A long-standing folk view is that static PTX metrics (register count, spills to local memory, instruction count) are leading indicators of measured GPU performance and can therefore gate CI without hardware execution. We test the claim on a controlled corpus.  \nThe claim is a bounded one. Static PTX metrics track structural changes to the kernel’s compiled work envelope. In this corpus, structural bugs produce nonzero ∆regs or ∆instrs, while semantic-only constant changes produce zero static delta. Measured runtime deltas are a different signal: at the sub-millisecond scale of these kernels, CUDA-event timing is dominated by launch and host variance, so measured ∆perf% does not reliably separate the two bug classes without larger shapes or per-architecture calibration. We give a measured example  \n2 D. Sarkar  \nof each. The cross-architecture sweep in Section 4.1 strengthens the portability side of the static claim. The static signal is identical across five GPU classes for the same kernel because it is determined at compile time.  \n2 Related Work  \nGPU performance modeling. The literature on GPU performance prediction from compiler-level information is rich [1,2] . The dominant approach uses a microarchitectural model (occupancy, memory bandwidth, instruction mix) to predict runtime. Correlation with measured performance is typically used to validate the model.  \nRegister pressure and spills. When register pressure exceeds the SM’sper-thread limit, ptxas spills to local memory backed by global DRAM [3,4] . This causes correctness-preserving but performance-destroying load and store traffic. Static spill detection through ld .local and st .local patterns in PTX is the typical CI gate.  \nTriton-level optimisation. Triton [6] compiles Python-level kernel descriptions to PTX with autotuning over BLOCK M, BLOCK N, and num warps. Librarylevel wrappers further trade off register pressure against tiling.  \nThe gap. No prior work, to our knowledge, controls for kernel semantics while varying static PTX metrics, and measures the regression-prediction correlation as a function of bug class. We do.  \n3 Method  \n3.1 Static metrics  \ncrates/gpuemu-daemon/src/artifact .rs parses PTX text and reports five metrics.  \n– register   count: total number of declared registers across all bank types. ","cbCain6nzC9ZF0Wr","https://ap.wps.com/l/cbCain6nzC9ZF0Wr","pdf",288935,1,9,"English","en",105,"# Introduction\n# Related Work\n# Method\n## Static metrics\n## Measured perf\n## The pairing protocol","[{\"question\":\"What static signals are compared in the study?\",\"answer\":\"The analysis extracts register count, spill count (via ld.local and st.local patterns), local memory bytes, and instruction count from emitted PTX, along with tracked instruction-pattern violations.\"},{\"question\":\"How do structural and semantic bugs behave differently according to the paper?\",\"answer\":\"Structural bugs that change the kernel’s compiled work envelope produce nonzero deltas in static PTX metrics, while semantic-only constant swaps can be invisible to static deltas (e.g., byte-identical PTX).\"},{\"question\":\"Why are measured runtime deltas not treated as a reliable gate here?\",\"answer\":\"At the sub-millisecond scale of the corpus kernels, CUDA-event timing is dominated by launch and host variance, so measured performance deltas are hardware- and noise-sensitive and do not consistently separate bug classes without further calibration.\"}]",1784182211,23,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"static-ptx-metrics-track-structural-kernel-regressions-but-miss-semantic-ones","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/static-ptx-metrics-track-structural-kernel-regressions-but-miss-semantic-ones/82673/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What static signals are compared in the study?","Question",{"text":75,"@type":76},"The analysis extracts register count, spill count (via ld.local and st.local patterns), local memory bytes, and instruction count from emitted PTX, along with tracked instruction-pattern violations.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How do structural and semantic bugs behave differently according to the paper?",{"text":80,"@type":76},"Structural bugs that change the kernel’s compiled work envelope produce nonzero deltas in static PTX metrics, while semantic-only constant swaps can be invisible to static deltas (e.g., byte-identical PTX).",{"name":82,"@type":73,"acceptedAnswer":83},"Why are measured runtime deltas not treated as a reliable gate here?",{"text":84,"@type":76},"At the sub-millisecond scale of the corpus kernels, CUDA-event timing is dominated by launch and host variance, so measured performance deltas are hardware- and noise-sensitive and do not consistently separate bug classes without further calibration.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":21,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]