stzGpu

(sans commentaire dans la source : le nom se lit comme une phrase) initaucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) is availableaucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) device nameaucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) adapter countaucune extension

Upload a vector; it lives on the device until Free() (or eviction under VRAM pressure -- bounded, counted, never silent).aucune extension

One-shot convenience: upload the bound vectors, run the kernel, read the result back, free everything temporary. Data in, DATA out.QCSXTZZZIBWU

Multi-input kernels over ALREADY-RESIDENT buffers: no transfer at all. paBindings binds vector names to stzGpuBuffer objects (plus scalars): oG.ApplyOnQ(k, [ :A = oBuf1, :B = oBuf2, :alpha = 2 ])aucune extension

Measure where the GPU actually starts winning ON THIS MACHINE, store the crossover in the calibration store, and persist it (the faces auto-load it in later sessions). Warm-min discipline throughout -- G0's clock inversion means sustained numbers flatter the GPU. Returns [ nThreshold, [ [nProblemSize, nCpuMs, nGpuMs], ... ] ].aucune extension

The same, on a caller-chosen ladder of corpus sizes (d fixed at 64: a representative embedding width; problem size = n*d).aucune extension

pCand may be an stzKernelMaker on the SAME declarations, or raw WGSL on the same binding contract (how a hand-written variant -- or a cheat -- gets checked). Options: [ :reps = 7, :band = 0.000001 ].aucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) verify withaucune extension

The roofline alone: could a dispatch touching nBytes finish in nMs on this device? Pure judgement over the MEASURED floors -- exposed so a guard can hand it an impossible number and watch it refuse.aucune extension

The measured floors themselves: [ :floorgbs, :submitfloorms ]aucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) load calibrationaucune extension

one rung: warm-min per-query ms through the REAL seam, both routes Measured 2026-09-09 on this laptop's RTX 3050: after ~10 s idle, every submit pays a fixed ~2 ms and the copy floor reads 13 GB/s instead of 85 -- and hundreds of tiny queries never lift it, while ~150 ms of heavy work does. Wake() is that work, ADAPTIVE: it copies until a copy runs at full speed, settles, and stops -- the budget is a cap. Every timing this face takes calls it first; a caller measuring on its own should too. Returns the copies it took (an awake device answers in ~20). THE CAUSE AND THE SYMPTOM. A shader that fails validation raises two errors: the parser's, with the offending name and line, and then the pipeline's, which only says the module is invalid. FirstError() is the parser's; LastError() is whatever came last. A compile clears both first.aucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) last erroraucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) clear errorsaucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) wakeaucune extension

Measured 2026-09-09 on the RTX 3050 laptop at 384 dims: the GPU loses at 1,000 (0.39x) and 4,000 (0.98x) and wins at 16,000 (1.42x); the multicore top-k switches on at n*d >= 8M and the window may close again above it -- which is why the rung above that gate is measured too, and why the store is SHAPED: a class that wins carries its win, and the flat line stays beyond the ladder where a class lost.aucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) calibrate knn resident withaucune extension

This is the ONE-SHOT doorway: upload, run, download. A plane holding its signal on the device uses the buffer form (StzEngineGpuOpConvolveReal over buffer ids) and pays no marshalling -- the sound desk's route.aucune extension

Report: [ :m, :n, :d, :hiddenn, :clocks, :refgpums, :refwallms, :variants = [ [name, verified, gpums, wallms, ratio, ratiowall], ... ], :winner = name, :ratio = generic/winner on the GPU clock, :stored = bool ]aucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) foundry pairdist withaucune extension

THE ATTENTION FOUNDRY (GK2c). The backbone's fused attention -- one workgroup per (head, query row) -- against three reduction-form variants at workgroup widths 64, 128 and 256, under the checker on the neural DLL's device; the winner recorded for the (tokens, width, head_dim) class under the op name "attention", persisted, and pushed into the neural DLL's table so the next forward dispatches it.aucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) foundry attention withaucune extension

THE MATMUL FOUNDRY (GK2b's matmul leg). C(m x n) = A(m x k) B(k x n): tile16 (the generic), tile8, reg2, reg4 under the checker at the asked shape and a hidden one; the winner recorded for the (m, n, k) class under the op name "matmul" and persisted. The backbone compiles the table's winner for its own shapes with the bias fused.aucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) foundry matmul withaucune extension

THE k-NN FOUNDRY (GS6e). The UMAP k-NN kernel lives in the stats DLL on its own device, so its enumeration runs there; the verdict is recorded in that DLL's variant table AND persisted here under the op name "umap_knn", beside pairdist's rows, so the next session finds it. The stats DLL reads the persisted rows back through StzUmapKnnVariantsSync() at its first fit.aucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) foundry umap knn withaucune extension

The kill line, per adapter: a grid of single-query shapes; TRUE if any class found a variant worth the margin.aucune extension

Report: [ :cells = [[d, n, cpuMs, gpuMs, ratio], ...], :flat = the flat line stored, :flatwhy = "crossover" | "beyond-ladder", :hidden = [ [d, n, cpuMs, gpuMs, predicted, measured, agree], ... ], :adapter = name ]aucune extension

(sans commentaire dans la source : le nom se lit comme une phrase) calibrate shaped withaucune extension