Research Papers
Research Papers news — Wednesday 07 October 2026
39 stories from 2 sources, every one in English. Each links to its publisher.
-
6G NeXt - Towards 6G Split Computing Network Applications: Use Cases and Architecture
arXiv cs.DC (Distributed)
arXiv:2610.08352v1 Announce Type: cross Abstract: The definition of the sixth generation mobile communication is in the full swing. Mobile communication aims for the convergence of physical, human and digital world. Research project 6G…
-
Cascadia: Resident 975B MoE Inference on Eleven AI PCs
arXiv cs.DC (Distributed)
arXiv:2610.07219v1 Announce Type: cross Abstract: Mixture-of-experts models make nearly trillion-parameter capacity accessible with sparse per-token computation, provided that the serving system can distribute the weights and coordinate…
-
Ofan: Optimal Load Balancing for AI Training
arXiv cs.DC (Distributed)
arXiv:2610.07230v1 Announce Type: cross Abstract: The extreme collective completion time (CCT) demands of AI workloads challenge existing packet spraying algorithms, which can have trouble efficiently load-balancing workloads that are sent…
-
NeMo-DCR: Bit-Exact Delta-Compressed Refit for Scalable Agentic RL at Trillion-Parameter Scale
arXiv cs.DC (Distributed)
arXiv:2610.08430v1 Announce Type: new Abstract: Agentic reinforcement learning (RL) disaggregates training from rollout, so each policy update must reach the rollout clusters before the next batch. Transferring a full 1T checkpoint for…
-
CUDA-MPQS: A GPU-Resident Self-Initializing Quadratic Sieve, and the Factorization of RSA-155
arXiv cs.DC (Distributed)
arXiv:2610.07126v1 Announce Type: cross Abstract: The quadratic sieve (QS) is an irregular, cache-hostile, branch-heavy integer factorization algorithm; prior GPU work accelerates individual stages of it. We present \swname{}, a…
-
A Resilient Runtime-Verification Fabric for Security Monitoring of Critical Edge-IoT Infrastructure
arXiv cs.DC (Distributed)
arXiv:2610.07282v1 Announce Type: cross Abstract: Protecting critical infrastructure increasingly depends on continuously verifying large IoT fleets against formal security specifications at runtime. Yet the runtime-verification (RV)…
-
LLM-Based Multi-Agent Collaboration for Constrained Multi-Objective Container Placement
arXiv cs.DC (Distributed)
arXiv:2610.07030v1 Announce Type: new Abstract: Container placement in data centers must simultaneously minimize power consumption and maximize affinity preferences, while satisfying multi-resource capacity and anti-affinity constraints…
-
DySCo: Dynamic Sharding for Collaborative Edge-Cloud LLM Inference with Depth-Synchronized Batching
arXiv cs.DC (Distributed)
arXiv:2610.08268v1 Announce Type: new Abstract: Pervasive intelligent applications are increasingly deployed on mobile and Internet of Things (IoT) edge devices. Consequently, Large Language Models (LLMs) are increasingly used to support…
-
NCCL M2N: A Layout- and Topology-Aware Collective for Distributed Tensor Resharding
arXiv cs.DC (Distributed)
arXiv:2610.07516v1 Announce Type: new Abstract: Distributed training and rollout generation often use different tensor layouts, requiring model weights to be resharded across distinct process groups. This M-to-N redistribution is not…
-
vTen: Tensor-Centric Verification Framework for Domain-Specific Accelerators
arXiv cs.DC (Distributed)
arXiv:2610.08372v1 Announce Type: new Abstract: The semantic gap between tensor-centric software models and signal-level hardware testbenches creates significant productivity bottlenecks in verifying Domain-Specific Accelerators (DSAs)…
-
KernelOPT: Dispatch-Aware Agentic Search for GPU Kernel Optimization
arXiv cs.DC (Distributed)
arXiv:2609.30059v2 Announce Type: replace Abstract: Deep learning inference and training performance depends critically on GPU kernel efficiency. Modern compilers such as PyTorch Inductor automatically generate GPU kernels from high-level…
-
Mosaic: GPU Sharing with Latency Guarantees through Kernel-Level Interference Prediction
arXiv cs.DC (Distributed)
arXiv:2610.07504v1 Announce Type: new Abstract: GPUs are increasingly in demand for AI workloads, yet often remain substantially underutilized, motivating workload colocation. However, colocation introduces interference that can degrade…
-
A Private IPFS Data Sanctuary for Verifiable Digital Collection Objects
arXiv cs.DC (Distributed)
arXiv:2610.07970v1 Announce Type: new Abstract: Background: Galleries, libraries, archives and museums (GLAM) increasingly produce scientific collection data whose digital state must remain identifiable, recoverable, and verifiable after…
-
Contextual Chain: Lightweight Continuity Authentication for Intermittently Connected Devices
arXiv cs.DC (Distributed)
arXiv:2610.08262v1 Announce Type: cross Abstract: Can authentication make memory, rather than computational hardness, the attacker's bottleneck? Contextual Chain is a lightweight continuity protocol for intermittently connected devices…
-
Productivity in HPC: A Critical Review from SLOC and Performance Portability to AI-Assisted Programming
arXiv cs.DC (Distributed)
arXiv:2610.07495v1 Announce Type: new Abstract: Productivity has long been recognized as a central concern in high-performance computing (HPC), yet it still lacks a workable formal definition and a broadly accepted metric for quantified…
-
Beyond Marginal Monitoring: Distributed Joint-Distribution Testing for Data Concept Drift in Large Scale E-Commerce Operations
arXiv cs.DC (Distributed)
arXiv:2610.08132v1 Announce Type: new Abstract: Concept drift threatens production machine learning, yet the empirical behavior of multivariate two-sample drift detectors at scale remains under-characterized. Existing benchmarks rarely…
-
Lachesis: Lifetime-Aware KV Cache Placement for Agent Serving across HBM and High-Bandwidth Flash
arXiv cs.DC (Distributed)
arXiv:2610.08378v1 Announce Type: cross Abstract: Large language model (LLM) serving is increasingly dominated by agentic workloads, in which agents and their sub-agents accumulate context as KV cache across many requests, consuming…
-
TRANSIT: Transparent Scale-in for Multi-Node LLM Training
arXiv cs.DC (Distributed)
arXiv:2610.07593v1 Announce Type: new Abstract: TRANSIT is a transparent scale-in framework to enable multi-node model training on fewer GPUs while maintaining training efficiency by transparently leveraging CPU DRAM as an extension of GPU…
-
T-CCL: Resource Efficient and Performant Collective Communication using Tensor Memory Accelerator
arXiv cs.DC (Distributed)
arXiv:2610.07098v1 Announce Type: new Abstract: Large transformer-based models increasingly depend on multi-GPU execution, which requires frequent collective communication among GPUs. Existing communication libraries often rely on many GPU…
-
Repository-Scale Performance Characterization of the IO500 Benchmark
arXiv cs.DC (Distributed)
arXiv:2610.07035v1 Announce Type: new Abstract: The IO500 benchmark provides a common basis for evaluating high-performance computing (HPC) storage systems, while its growing submission repository also offers an opportunity to study…
-
Evaluating Behavioral Context for Interpretable IAM Policy Risk Scoring in Cloud Environments
arXiv cs.DC (Distributed)
arXiv:2610.07345v1 Announce Type: cross Abstract: IAM policy analysis typically emphasizes the authorization capabilities encoded in a policy, but security analyst review priority may also depend on the behavioral and environmental context…
-
When Can Stateless Recovery Defeat Byzantine Quorum Safety? A Tight Normal Form for Single-Step BFT
arXiv cs.DC (Distributed)
arXiv:2610.07759v1 Announce Type: new Abstract: Byzantine quorum safety relies on correct replicas refusing to sign conflicting values. A replica that loses its protocol state during recovery but retains its identity and signing key may…
-
Memory-Efficient Expert Routing for Distributed MoE Training
arXiv cs.DC (Distributed)
arXiv:2610.07333v1 Announce Type: new Abstract: As Mixture-of-Experts (MoE) models scale toward hundreds of experts and higher top-$k$ routing, memory efficiency in distributed training becomes a critical bottleneck. Peak memory is…
-
Do I Need the Cloud? Uncertainty-Aware Step-Level Handoff for Small Language Model Agents
arXiv cs.DC (Distributed)
arXiv:2610.07816v1 Announce Type: cross Abstract: Small language models (SLMs) are attractive as local agent controllers because they reduce remote inference, latency, and deployment footprint, yet structured tool errors can cause an agent…
-
GPU Acceleration of Awkward Arrays: Using Python cuda.compute
arXiv cs.DC (Distributed)
arXiv:2610.08238v1 Announce Type: new Abstract: Awkward Array is a widely used library in high-energy physics (HEP) for representing and manipulating nested, variable-length data in Python. Previous CHEP contributions have explored GPU…
-
Tram-FL: Reducing Communication and Computation Costs through Sequential Model Circulation in Decentralized Federated Learning
arXiv cs.DC (Distributed)
arXiv:2610.07859v1 Announce Type: cross Abstract: Conventional decentralized federated learning (DFL) often focuses on clients, with each client maintaining a model copy, performing updates individually, and undertaking model exchange and…
-
SketchSSM: Write to the Full State, Read from a Compact Sketch
arXiv cs.DC (Distributed)
arXiv:2609.33051v2 Announce Type: replace-cross Abstract: Hybrid-attention models replace most softmax attention layers with linear attention, reducing KV-cache growth and enabling larger decode batches where recurrent-state access becomes…
-
When Robots Crash: Optimal Asynchronous Gathering at Weber Meeting Nodes
arXiv cs.DC (Distributed)
arXiv:2610.06939v1 Announce Type: new Abstract: We study the \textit{optimal gathering} problem over a finite set of designated \textit{meeting nodes} for \textit{asynchronous, anonymous,} and \textit{oblivious} mobile robots on an…
-
One Global Beam Across Many GPUs: High-Throughput Beam Search at Billion-Record Frontier Scale
arXiv cs.DC (Distributed)
arXiv:2610.06718v2 Announce Type: replace Abstract: Beam search repeatedly makes many children, removes duplicates, and keeps the best $B$. We show how many GPUs can perform these steps as one search even when the retained set does not fit…
-
AID: A Framework for AI Infrastructure Dynamics
arXiv cs.DC (Distributed)
arXiv:2610.04801v2 Announce Type: replace Abstract: A useful model of AI inference infrastructure must specify the system state, the information available to an observer, and the decisions the model is intended to support. We introduce AID…
-
AEGIS: Runtime-Guided GPU Collocation for Multi-Tenant Deep Learning Training
arXiv cs.DC (Distributed)
arXiv:2508.19073v4 Announce Type: replace Abstract: Deep learning training commonly runs on shared multi-tenant GPU servers, where exclusive allocation provides isolation but can leave resources underutilized and increase queueing time…
-
FailBench: Evaluating Fault Tolerance Across Distributed Training Architectures
arXiv cs.DC (Distributed)
arXiv:2610.07688v1 Announce Type: new Abstract: Distributed deep learning relies on data, pipeline, tensor, and hybrid parallelism, yet fault-tolerance mechanisms are typically evaluated only on the architecture for which they were…
-
Evaluating Inference Compute for Generative AI: A Framework for Enterprise Workloads
arXiv cs.DC (Distributed)
arXiv:2610.07094v1 Announce Type: cross Abstract: LLM deployment is shifting from single-turn completion to agentic trajectories in which a model plans, calls tools, reads results and reasons at test time before acting. This inverts the…
-
Persistent Memory in Multi-Agent LLM Inference: What It Costs, What It Buys, and When You Can Tell
arXiv cs.DC (Distributed)
arXiv:2610.07782v1 Announce Type: cross Abstract: Decomposing long-context inference across cooperating agents bounds the active KV cache per call rather than total evidence, which matters when KV-cache memory binds. Many such systems add…
-
High Speed Network Processing for Scientific Computing in the Open Cloud Testbed
arXiv cs.DC (Distributed)
arXiv:2610.07377v1 Announce Type: cross Abstract: This paper presents an FPGA-based SmartNIC architecture designed to address critical network challenges in data-intensive scientific computing. As scientific workflows increasingly demand…
-
Terracotta: Enabling the Adoption of New DRAM Techniques via a Flexible DRAM Interface and Memory Controller
arXiv cs.DC (Distributed)
arXiv:2610.06475v2 Announce Type: replace-cross Abstract: DRAM continues to limit the performance, energy efficiency, and robustness of modern systems. Many prior works propose DRAM techniques that support in-DRAM computation, improve…
-
Byzantine-Tolerant Causal Unicast with Constant Message Space Overhead
arXiv cs.DC (Distributed)
arXiv:2610.07368v1 Announce Type: new Abstract: Causal message ordering provides essential semantics for distributed applications, yet ensuring it within an asynchronous system subject to Byzantine failures presents fundamental theoretical…
-
Why we must advocate for proper care on the frontlines of the Ebola outbreak
Nature Technology
-
Characterizing the role of reasoning tokens in large language models as state over tokens
Nature Technology