Agent skill · software engineering · nvidia
tilegym-converting-cutile-to-triton
Converts cuTile GPU kernels (@ct.kernel) to Triton (@triton.jit). Handles standard in-repo conversion, debugging (cudaErrorIllegalAddress, shape mismatch, numerical mismatch), and mapping cuTile idioms (ct.load/ct.store, ct.Constant, ct.launch) to Triton equivalents. Covers dual-kernel layout flags (e.g. transpose=True/False + autotune grid via META) per translations/advanced-patterns.md. Use when converting, porting, or translating cuTile kernels to Triton, or debugging existing Triton translations.
Why this skill is useful
Adds a detailed executable workflow for converting cuTile kernels to Triton, including specific debugging and optimization strategies not commonly found in public documentation.
What it needs
Requires bash installed locally. About 8k tokens when loaded. Last updated 2026-08-07. 2,816 stars on the source repository.
What this skill does
cuTile → Triton Conversion Convert @ct.kernel kernels to @triton.jit. API mapping: references/api-mapping.md (cuTile → Triton). In this skill’s Markdown, Triton launch syntax kernel［grid］(…) uses Unicode brackets so link checkers do not parse grid as a hyperlink; use normal ASCII brackets in real Triton code. Instructions Follow the phase-gated workflow in translations/workflow.md. Every conversion should go through analyze → convert → validate → test → benchmark, with explicit gates before moving on. Use the documents in Workflow Selection when the task matches a special case (errors, layout flags, perf). 0. Optimization strategy (perf-sensitive / attention) — If the op is attention, FMHA, sliding window, soft cap, or GQA (e.g. Gemma gemmaattention), read references/optimization-strategy.md before converting the inner loop, then apply §4 Gemma FMHA checklist. For other GEMM/BMM/attention-adjacent kernels, still skim §2–§3 of that file after TMA is done. 1. Select path — Existing TileGym op: standard mode in translations/workflow.md. If the cuTile source uses transpose / transposev, dual layouts, or MLA-style paths, read translations/advanced-patterns.md before writing Triton (two kernels + META grid, not one kernel + tl.trans). 2. Pre-flight — Run the Pre-flight Analysis grep commands on the cuTile source. Count @ct.kernel definitions; note TMA-relevant ct.load/ct.store, ct.launch, Constant, and layout flags. 3. Read mapping — Keep references/api-mapping.md open for cuTile → Triton API pairs. For runtime failures (illegal address, dtype, strides), use references/debugging.md. 4. Convert — Copy the Conversion Checklist into a todo list and execute in order. Structure and file placement: translations/file-structure.md. Mandatory: any 2D+ block-shaped tile load/store uses tl.maketensordescriptor (TMA), not raw tl.load(ptr+offs, mask=…) for full tiles—skipping this is the most common source of large regressions. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills nvidia/tilegym-converting-cutile-to-triton