aie-dataflow-opt

aie-dataflow-opt is an MLOps & model serving skill for AI agents, published by Xilinx in the mlir-aie repository on GitHub. Skillspace rates aie-dataflow-opt C (74/100) and it passed the safety scan; the rating is based on static analysis only.

Guide to optimizing an AIE/IRON design's dataflow — dispatch partitioning, tile placement, overlays, weight/activation streaming strategy, and DMA bandwidth/compression — once it's already functionally correct. Distinct from aie-kernel-opt (which optimizes a single compiled kernel in place): this is about which op runs on which tile or dispatch, how data moves between them, and where the real bottleneck is before touching any kernel at all. Use this whenever the user is optimizing throughput/latency of a full IRON design (not a single kernel), deciding tile placement, overlay layout, or dispatch boundaries for a multi-dispatch model, choosing between static/streamed weights, modeling DMA bandwidth or compression ratios, or asking why a design is slow when profiling shows no single kernel is unusually expensive — even if they haven't used the word "dataflow.

Grade

C

Quality

74 / 100

Safety

Passed

GitHub stars

687

Unproven: rated by static analysis only, not yet compared head-to-head.

Our two independent classifiers differ on this skill's category, so treat the placement as provisional.

This rating describes the version of the repository we crawled; it may have changed since, so go by its current files.

Skillspace indexes and rates this skill; it does not host or endorse it. Run third-party skills at your own risk.