Metraipolis › archive › dc_8adbc80803dd

Execution-Path Qualification and Realized Costs in Speculative Decoding on Consumer Systems — arXiv:2610.03228

Source: arXiv
Original: https://arxiv.org/abs/2610.03228
License (as archived): CC0-1.0
Archived: 2026-10-05T22:54:44.417Z
SHA-256 of the archived text: a99fd4a4bb92ab52df727dc44983e70b873032ab86c580a571060c848d639c05
Stable URL of this page: https://metraipolis.com/documento/dc_8adbc80803dd

How to cite

Execution-Path Qualification and Realized Costs in Speculative Decoding on Consumer Systems — arXiv:2610.03228 — arXiv, https://arxiv.org/abs/2610.03228 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8adbc80803dd (Metraipolis archive dc_8adbc80803dd, sha256 a99fd4a4bb92ab52df727dc44983e70b873032ab86c580a571060c848d639c05)

BibTeX

@misc{dc8adbc80803dd,
  title = {Execution-Path Qualification and Realized Costs in Speculative Decoding on Consumer Systems — arXiv:2610.03228},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.03228},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8adbc80803dd; sha256 a99fd4a4bb92ab52df727dc44983e70b873032ab86c580a571060c848d639c05},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.03228. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Numerical execution paths can change greedy output; a mismatch alone does not predict the remaining trajectory. On the tested B570 stack, we select a draft-free target-only micro-batch-one predictor using exposed discovery data, then freeze complete generated-ID predictions before reserved live speculation. It matches 72/72 capped sequences on 24 repeat-stable prompts, including all 12 runs of four reference-divergent prompts; the same-set reference matches 60/72. The predictor is sufficient for this path, but it neither restores qualification nor identifies a unique numerical cause. Separate B570 same-state forks test intervention effects; RTX proposal/replay programs test reproduction while sharing numerical routines. We assess cost separately. Newly direct-ID-qualified RTX serialization is slower within its instrumented post-prefill boundary ($0.646\times$ target/serialized latency). At fixed measured progress and target work, 358H remains below parity even with draft and residual costs removed. Qualified M4 Guard beats target-only but loses to fixed speculation. These results give testable diagnoses and engineering decisions about lossless compatibility and complete cost on the measured paths; they establish no portable decoding or control gain.

Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.PF, nuovi lavori del giorno).
Categorie arXiv: Performance (cs.PF).
Pagina del lavoro: https://arxiv.org/abs/2610.03228 — 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 2026-10-05T22:46:55.523Z dalla pagina https://arxiv.org/list/cs.PF/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).