arXiv preprint ·
An Information-Geometric Framework for OCS-Reconfigurable Interconnect Topology in Large-Scale AI Clusters
- Published
- Source
- arXiv preprint, posted before peer review
- Original
- Read it at arXiv
- Topics
Abstract
Optical Circuit Switches (OCS) make inter-pod wiring in AI clusters programmable, posing design questions on wiring choice, symmetry departure, and reconfigurations. We present a unified information-geometric framework to address these issues.A wiring is a point in a probability simplex over circuits within a degree polytope. Departure from a symmetric Fat-Tree prior $\mathbf{q}$ is measured by $\DKL(\cdot \,\|\, \mathbf{q})$. The cost $U$ is the smoothed DAG critical-path length, whose gradient equals path marginals of a Gibbs distribution via forward--backward recursion.For free energy $F_{β_1} = U + β_1^{-1} \DKL(\cdot \,\|\, \mathbf{q})$, we prove well-posedness, monotonic dependence on $β_1$, adaptation bounds via Jeffreys divergence, and a symmetry principle for static optimality. An entropic proximal-gradient method updates multiplicatively with a $|V|$-dim dual KL projection, achieving linear convergence.For time-varying traffic, a threshold policy yields regret $O(\sqrt{ΓC V_1 T})$ against the switching-free optimum. Traffic changes are detected by a multinomial GLR--CUSUM detector with non-asymptotic false-alarm bounds. Multi-job generalized-mean fairness aligns with $α$-fairness ($α\ge 2$). Experiments on pipeline, data, and MoE workloads validate the framework against integer programming.
The authors’ abstract, as arXiv lists it. arXiv’s listings, abstracts included, are free for anyone to reuse.
Read it on ReadSearch
Each opens after you create an account, and brings you back to this paper.
More in Distributed & Parallel Computing
Conjugate Compilation: Self-Improving Agentic Co-Compilation for Speculative Decoding
Peer-Relative Trust-Guided Update Screening for Robust Semi-Decentralized Federated Learning under Label-Flipping Attacks
Bidirectionality is Necessary in Cellular Machines for Exact-Arrival Multi-Agent Path Finding
DROD: Dual-Ratio Ordered Dropout for Federated Continual Learning on Heterogeneous Devices
A Generic On-Board AI Processing Chain for Earth Observation: From Application Development to Embedded Deployment and Mission-Level Orchestration
Get the few papers worth your time in Distributed & Parallel Computing every morning.
ReadSearch reads the new papers in your field each day and brings you the few worth reading, each with its key finding in a sentence.