[{"data":1,"prerenderedAt":174},["ShallowReactive",2],{"run:c1-drums-12500-off-prof":3,"run-findings:c1-drums-12500-off-prof":153},{"run":4,"config":18,"metrics":98,"curve":104,"assets":152},{"id":5,"title":5,"subtitle":6,"eyebrow":7,"date":8,"workspace":9,"commit":10,"tags":11,"verdict":17},"c1-drums-12500-off-prof","drums (998k splats) per-kernel gpu_ms share 計測。Lego v2 (p3-kernel-share-post-s3 §6) の測定 regime を drums dataset で再現し、splat 数 regime の違いが share 構造に影響するかを判定する","Run summary · 2026-09-seminar05-milestones (次フェーズ タスクリスト — 多 scene per-kernel share 検証)","2026-08-31","splat","3db33a8",[12,13,14,15,16],"p3-kernel-share-drums","multi-scene","drums","per-kernel-gpu-ms","gpu-gate","partial",[19,22,25,28,31,34,37,40,43,46,49,52,55,58,61,64,67,70,73,76,79,81,84,87,89,92,95],{"key":20,"value":21},"dataset","\u002FUsers\u002Fotkrickey\u002Fdev\u002F3dgs-workspace\u002Fdatasets\u002Fnerf_synthetic\u002Fdrums",{"key":23,"value":24},"gt_convention","Premultiplied",{"key":26,"value":27},"iterations","12,500",{"key":29,"value":30},"seed","42",{"key":32,"value":33},"capacity","1,000,000 splats",{"key":35,"value":36},"sh_degree","3",{"key":38,"value":39},"loss","L1Ssim",{"key":41,"value":42},"lambda","0.200",{"key":44,"value":45},"ssim","window=7 sigma=1",{"key":47,"value":48},"backend.backward","Simd",{"key":50,"value":51},"backend.loss_path","Gpu",{"key":53,"value":54},"backend.emit_pairs","Original",{"key":56,"value":57},"backend.forward","F32",{"key":59,"value":60},"backend.dispatch","PerStage",{"key":62,"value":63},"backend.radix_scan","GpuHierarchical",{"key":65,"value":66},"backend.radix_scatter","SimdBallot",{"key":68,"value":69},"backend.binning","Resident",{"key":71,"value":72},"backend.tile_sort","GlobalRadix",{"key":74,"value":75},"backend.submission","PerKernel",{"key":77,"value":78},"backend.target_cache","true",{"key":80,"value":78},"backend.shader_fast_math",{"key":82,"value":83},"os_build","26.6.2 (25G83)",{"key":85,"value":86},"backend.ssim_clear","HostMemset",{"key":88,"value":86},"backend.raster_grad_clear",{"key":90,"value":91},"backend.backward_uniform","HostBuffer",{"key":93,"value":94},"backend.lazy_c3_scratch","false",{"key":96,"value":97},"backend.ssim_walk","PerChannel",{"psnr":99,"wallclock":100,"wallclock_regress":101,"splats":102,"final_loss":103},24.517929077148438,"4m 9s",false,748708,"1.765736e-2",{"loss":105,"splats":136},{"iters":106,"values":121},[107,108,109,110,111,112,113,114,115,116,117,118,119,120],1,1000,2000,3000,4000,5000,6000,7000,8000,9000,10000,11000,12000,12500,[122,123,124,125,126,127,128,129,130,131,132,133,134,135],0.6063743829727173,0.0507052019238472,0.03167391195893288,0.028153004124760628,0.025068026036024094,0.022766783833503723,0.01991906948387623,0.018875960260629654,0.018197443336248398,0.017932618036866188,0.017838727682828903,0.017586704343557358,0.017527448013424873,0.017657358199357986,{"iters":137,"values":138},[107,108,109,110,111,112,113,114,115,116,117,118,119,120],[139,140,141,142,143,144,145,146,147,148,149,150,151,102],5207,5606,122843,265921,381505,464360,529695,581745,619505,650448,678077,705146,733488,{"has_renders":101,"has_splat":101},[154],{"id":12,"title":155,"date":8,"status":156,"polarity":157,"category":158,"axes":159,"tags":161,"task_code":164,"related_runs":165,"delta_psnr":168,"delta_wallclock":169,"rank":170,"verdict":171,"impact_summary":172,"detail_path":173},"drums (998k splats @30k, in-window 749k @12.5k) per-kernel share 実測 — share 構造は Lego とほぼ不変 (backward_raster 40.09% vs 40.07%)、絶対 ms\u002Fiter 差は binary vintage (S4b+C2) で汚染","stable","mixed","measurement",[107,160],3,[12,13,14,15,162,16,163],"kernel-share","binary-vintage-caveat","次フェーズ タスクリスト — 多 scene 検証 (drums per-kernel share)",[5,166,167],"c1v2-12500-off-prof","c1v3-12500-off-prof","n\u002Fa (drums self-eval mean val PSNR 24.518 dB @12.5k — non-converged 途中経過、多 scene 退行チェックの正式値ではない。参照: multiscene-drums-30k baseline は 30k 収束時 27.217 dB)","drums: train elapsed 249.77s (n=1) \u002F Lego v2: 174.57s (n=1)。**両方とも production baseline とは比較しない** — submission=per_kernel\u002Fforward_submission=per_stage\u002Fpipelining=off の 3 重に非 production な測定専用 regime (busy% drums 68.3%、Lego v2 65.6%)。意味があるのは kernel 単体の gpu_ms とその share のみ","medium","share-structure-invariant-under-1.85x-splat-scale \u002F absolute-ms-comparison-contaminated-by-binary-vintage","Lego で判断された今日の最適化候補が drums (998k splats @30k、Lego の 2.5倍) でも同じ share 構造で有効かを判定した。drums 12.5k を Lego v2 と同一 regime (submission=per_kernel\u002Fforward_submission=per_stage\u002Fpipelining=off) で 1 run 実測した結果、\u003Cstrong>per-kernel share は Lego v2 とほぼ同一\u003C\u002Fstrong> (backward_raster 40.09% vs 40.07%、adam 15.52% vs 15.28%、rasterize 14.94% vs 14.13%、ssim 12.75% vs 13.12%、radix 11.37% vs 12.01% — 全 kernel で Δ≤1.5pt)。\u003Cstrong>splat 数が 1.85倍 (in-window 748,708 vs 404,652) 違っても share 順位・構造がほぼ動かない\u003C\u002Fstrong>という結果は、Lego で優先度付けした施策 (M3 系 backward・M4 radix・S4b ssim 等) の相対重要度が drums でも維持されることを示唆する。\u003Cstrong>ただし絶対 ms\u002Fiter の比較は binary vintage 差で汚染されている\u003C\u002Fstrong> — 比較対象の Lego v2 は commit a446f2c、drums は HEAD 3db33a8 で、間に S4b (ssim_reduce=gpu default flip) と C2 計装 (全 CB へ kernel timestamp 収集追加) が入っている。splat 数と無関係なはずの ssim_fwd_grad が 1.446倍 (1.202→1.738 ms\u002Fiter) 増えている一方、全 9 leaf kernel が揃って 1.17〜1.93倍という比較的均一な増加を示しており、これは\u003Cstrong>純粋な splat-count scaling では説明しにくく、binary vintage (S4b\u002FC2) 由来の系統的な底上げが乗っている可能性が高い\u003C\u002Fstrong>。よって「share の相対構造」は信頼できるが「ms\u002Fiter の絶対値・scaling 係数」は次ラウンドで Lego 側も HEAD で撮り直すまで確定できない。","\u002Ffindings\u002Fp3-kernel-share-drums\u002F",1790227323336]