agentleFS
onnxonnx/onnx-mlir/.agents/skills/nnpa-internals/SKILL.md

nnpa-internals

Skill1.1k starsChanged 12 days agoApache-2.0View on GitHub

Background on the NNPA (IBM Z AI accelerator) path in onnx-mlir: ztensor layouts and stickification, stick/unstick SIMD codegen, broadcasting rules, the build-time gate, placement and tuning flags, ONNX-level rewrites before ZHigh lowering, and NNPA op fusion. Use when you need to understand why an op does or doesn''t offload, or how NNPA data layouts work.

Trending

Quiet since December.about the whole repository

Dec 2025: 1 mentionJan 2026: 0 mentionsFeb 2026: 0 mentionsMar 2026: 0 mentionsApr 2026: 0 mentionsMay 2026: 0 mentionsJun 2026: 0 mentionsJul 2026: 0 mentionsAug 2026: 0 mentionsSep 2026: 0 mentionsOct 2026: 0 mentionsNov 2026: 0 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the onnx/onnx-mlir repository that our weekly crawl found. There is no adoption data for this file yet.

Who's talking about onnx/onnx-mlir

The biggest 5 of 22 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 22

Talked about on GitHub (22 times).

On GitHub · 22

17 more on GitHub

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read only the reference you need:

    line 12

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md
<!-- Generated from the onnx-mlir knowledge base (generated briefing), verified against onnx-mlir@b505d64b. Do not edit here: see .agents/skills/README.md. -->

NNPA Internals

Reference material that nnpa-performance and the other task skills in this directory link to. For per-op limits see docs/SupportedONNXOps-NNPA.md; for usage see docs/AccelNNPAHowToUseAndTest.md.

Read only the reference you need:

  • NNPA Broadcasting Support: Which NNPA operations tolerate input operands of different shapes, and what happens when they don't match exactly.
  • NNPA ZTensor Data Layouts: Every ZTensor (ZTensorEncodingAttr) carries a DataLayout value that determines how its logical dimensions are tiled into the physical 6D stick layout NNPA hardware operates on.
  • SIMD Code Generation for Stick/Unstick Operations: SIMD vectorized code for stickification and unstickification is generated by onnx-mlir during the ZHighToZLow lowering phase, using compile-time template instantiation to emit Krnl/Affine/SCF loop constructs that leverage vector operations.
  • NNPA Accelerator Path (selected sections): The NNPA (Neural Network Processing Assist) is IBM's hardware accelerator for neural network inference, integrated into the z16 Telum processor.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.