Rank Portability Does Not Imply Feasibility Portability: Target-Specific Evaluation of Joint Hardware Constraints

Abstract:Cross-device hardware evaluation often assumes that if architecture rankings transfer across devices, a proxy device can support target-side model selection. We stress-test this assumption for joint latency-energy feasibility across two public architecture families. On NAS-Bench-201, cross-device rank correlations are moderate, while target-comparable feasible-set overlap remains incomplete. A faithful AdaProxy diagnostic substantially improves latency ranking, showing that the observed boundary failures are not simply due to weak adaptation. Exact finite-sample split-conformal analysis also exposes an evidence bottleneck: a finite one-sided 90% threshold requires at least nine calibration observations. We then replicate the phenomenon on 10,000 GPT architectures across 13 HW-GPT-Bench devices. Relative to an RTX3080 proxy, target latency SRCC ranges from 0.951 to 0.996, yet proxy-reuse violation risk ranges from 33.3% to 100% under matched joint constraints. These results show that rank portability, feasibility portability, and target-specific decision support are distinct evaluation objects. Cross-device evaluations should therefore report which target environments actually support the operating point being claimed.

Machine Learning (cs.LG)

arXiv:2609.22122 [cs.LG]

(or arXiv:2609.22122v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2609.22122 Focus to learn more arXiv-issued DOI via DataCite

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.