Metraipolis › archive › dc_ccef40449b54

Schedules Are Solvable Symbols: Tuning-Free Compilation of Tile Programs on Dataflow Architectures — arXiv:2609.29219

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.29219
License (as archived)
CC0-1.0
Archived
2026-10-08T07:30:40.431Z
SHA-256 of the archived text
0f2051e7489844f5317d23fb9ca8a4e127a2593e9af5cf22780c294029dd42e8
Stable URL of this page
https://metraipolis.com/documento/dc_ccef40449b54

How to cite

Schedules Are Solvable Symbols: Tuning-Free Compilation of Tile Programs on Dataflow Architectures — arXiv:2609.29219 — arXiv (en), https://arxiv.org/abs/2609.29219 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ccef40449b54 (Metraipolis archive dc_ccef40449b54, sha256 0f2051e7489844f5317d23fb9ca8a4e127a2593e9af5cf22780c294029dd42e8)
BibTeX
@misc{dcccef40449b54,
  title = {Schedules Are Solvable Symbols: Tuning-Free Compilation of Tile Programs on Dataflow Architectures — arXiv:2609.29219},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.29219},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ccef40449b54; sha256 0f2051e7489844f5317d23fb9ca8a4e127a2593e9af5cf22780c294029dd42e8},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.29219. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Modern AI and HPC accelerators increasingly expose dataflow features: software-visible mechanisms for data movement and overlap, such as inter-core communication through the on-chip network and intra-core asynchronous pipelining. These features shift scheduling responsibility from hardware to the compiler, and because placement, movement, and synchronization become software-visible, they also make the performance of static schedules predictable. Yet high performance on such hardware still relies on vendor-engineered kernel libraries or profile-based auto-tuning, whose embedded expert knowledge transfers poorly across architectures and algorithms. We present Loom, a tuning-free symbolic compiler framework for tile-based SPMD programs on spatial dataflow architectures. The central idea is to treat tile-based SPMD compilation as a hardware-explicit static optimization problem. Loom enumerates discrete spatial-mapping and communication candidates while keeping value parameters, such as tiling factors and pipeline knobs, symbolic within each candidate. From an explicit hardware description, it derives symbolic legality constraints and latency expressions, formulates one CP-SAT problem per schedule candidate, and jointly solves inter-core dataflow, intra-core asynchronous scheduling, and block sizes at compile time. On two Tenstorrent generations, Wormhole and Blackhole, Loom matches or exceeds the vendor-optimized TTNN library on GEMM, Flash Attention, and Flash Decode, out of the box and without per-shape profiling or profile-based platform-specific schedule tuning. These results suggest that hardware-derived symbolic compilation provides a retargetable alternative to profiling-based tuning for spatial dataflow architectures while remaining interpretable by keeping optimization decisions traceable to source-level symbols.

Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-09-25 in cs.PL).
Categorie arXiv: Programming Languages (cs.PL).
Pagina del lavoro: https://arxiv.org/abs/2609.29219 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-08T06:45:28.396Z dalla pagina https://arxiv.org/catchup/cs.PL/2026-09-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).