Related papers: Managing Communication Latency-Hiding at Runtime f…
Despite advancements in the areas of parallel and distributed computing, the complexity of programming on High Performance Computing (HPC) resources has deterred many domain experts, especially in the areas of machine learning and…
Search strategies are crucial to efficiently solve constraint satisfaction problems. However, programming search strategies in the existing constraint solvers is a daunting task and constraint-based languages usually have compositionality…
Most multi-agent systems rely exclusively on autoregressive language models (ARMs) that are based on sequential generation. Although effective for fluent text, ARMs limit global reasoning and plan revision. On the other hand, Discrete…
The cost of communication is a substantial factor affecting the scalability of many distributed applications. Every message sent can incur a cost in storage, computation, energy and bandwidth. Consequently, reducing the communication costs…
We consider both leaderless and leader-follower, possibly nonlinear, networks affected by time-varying communication delays. For such systems, we give a set of sufficient conditions that guarantee the convergence of the network towards some…
The execution behavior of a program often depends on external resources, such as program inputs or file contents, and so cannot be run in isolation. Nevertheless, software developers benefit from fast iteration loops where automated tools…
Large language models (LLMs) inference is both expensive and slow. Local caching of responses offers a practical solution to reduce the cost and latency of LLM queries. In research contexts, caching also enhances reproducibility and…
Large-scale machine learning training, in particular distributed stochastic gradient descent, needs to be robust to inherent system variability such as node straggling and random communication delays. This work considers a distributed…
In this paper, we consider a hierarchical distributed multi-task learning (MTL) system where distributed users wish to jointly learn different models orchestrated by a central server with the help of a layer of multiple relays. Since the…
We study in this paper the impact of communication latency on the classical Work Stealing load balancing algorithm. Our paper extends the reference model in which we introduce a latency parameter. By using a theoretical analysis and…
We present Prompt Cache, an approach for accelerating inference for large language models (LLM) by reusing attention states across different LLM prompts. Many input prompts have overlapping text segments, such as system messages, prompt…
Large Language Models (LLMs) have pushed the frontier of artificial intelligence but are comprised of hundreds of billions of parameters and operations. For faster inference latency, LLMs are deployed on multiple hardware accelerators…
This paper focuses on the analysis of real-time non preemptive multiprocessor scheduling with precedence and several latency constraints. It aims to specify a schedulability condition which enables a designer to check a priori -without…
Context: The term reactivity is popular in two areas of research: programming languages and distributed systems. On one hand, reactive programming is a paradigm which provides programmers with the means to declaratively write event-driven…
Timing control while preserving determinism is often a key requirement for ensuring the safety and correctness of distributed cyber-physical systems (CPS). Discrete-event (DE) systems provide a suitable model of computation (MoC) for…
Fine-tuning is the process of adapting the pre-trained large language models (LLMs) for downstream tasks. Due to substantial parameters, fine-tuning LLMs on mobile devices demands considerable memory resources, and suffers from high…
Tool-augmented LLMs are increasingly deployed as agents that interleave natural-language reasoning with executable Python actions, as in CodeAct-style frameworks. In deployment, these agents rely on runtime state that persists across steps.…
The ever-increasing sizes of large language models necessitate distributed solutions for fast inference that exploit multi-dimensional parallelism, where computational loads are split across various accelerators such as GPU clusters.…
Shared resource interference is observed by applications as dynamic performance asymmetry. Prior art has developed approaches to reduce the impact of performance asymmetry mainly at the operating system and architectural levels. In this…
Parallel task-based programming models, like OpenMP, allow application developers to easily create a parallel version of their sequential codes. The standard OpenMP 4.0 introduced the possibility of describing a set of data dependences per…