SearcharxivSearch

arXiv · 2512.04705

Hardware-Algorithm Co-Optimization of Early-Exit Neural Networks for Multi-Core Edge Accelerators

Abstract

Deployment of dynamic neural networks on edge accelerators requires careful consideration of hardware constraints beyond conventional complexity metrics such as Multiply-Accumulate operations. In Early-Exiting Neural Networks (EENN), exit placement, quantization level, and hardware workload mapping interact in non-trivial ways, influencing memory traffic, accelerator utilization, and ultimately energy-latency trade-offs. These interactions remain insufficiently understood in existing Neural Architecture Search (NAS) approaches, which typically rely on proxy metrics or hardware-in-the-loop evaluation. This work presents a hardware-algorithm co-design framework for EENN that explicitly models the interplay between quantization, exit configuration, and multi-core accelerator mapping. Using analytical design space exploration, we characterize how small architectural variations can induce disproportionate changes in hardware efficiency due to tensor dimension alignment and dataflow effects. Building on this analysis, we formulate EENN deployment as a constrained multi-objective optimization problem balancing accuracy, energy-latency product, exit overhead, and dynamic inference behavior. Experimental results on CIFAR-10 demonstrate that the proposed framework identifies architectures achieving over 50\% reduction in energy-latency product compared to static baselines under 8-bit quantization. The results highlight the importance of deployment-aware co-design for dynamic inference on heterogeneous edge platforms.

Explore related subjects

Keep this discovery

BibTeXRIS

Alaa Zniber, Arne Symons, Ouassim Karrakchou, Marian Verhelst, Mounir Ghogho. 2025-12-04. Hardware-Algorithm Co-Optimization of Early-Exit Neural Networks for Multi-Core Edge Accelerators. https://arxiv.org/abs/2512.04705

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

The Computational Complexity of Holant Problems on 4-regular Graphs from the Stable Subgroup Sequence of $SL(2,\mathbb{C})$

The Holant framework provides a general setting for studying counting problems and includes graph homomorphisms (\#GH) and counting constraint satisfaction problems (\#CSP) as special cases. Over the past twenty years, a series of computational complexity dichotomies have been established for Holant problems, but the classification for complex-valued signatures is still open. The main obstacle is the case in which all signatures have even arity. In this paper, we establish a dichotomy for Holant problems with a complex-valued 4-ary signature, which is a key base case for the full classification of Holant problems. We present a new strategy by introducing Schur's theorem, the classification of finite subgroups of $\mathrm{SL}(2,\mathbb{C})$ and stable subgroup sequences into the proof. These new techniques are of independent interest.

cs.CC

Topology inside NC$^1$

We show that ACC$^0$ is precisely what can be computed with constant-width circuits of polynomial size and polylogarithmic genus. This extends a characterization given by Hansen, showing that planar constant-width circuits also characterize ACC$^0$. Thus polylogarithmic genus provides no additional computational power in this model. We consider other generalizations of planarity, including crossing number and thickness. We show that constant-width circuits of polynomial size and thickness two already suffice to capture all of NC$^1$.

cs.CC