Related papers: A Doubly-pipelined, Dual-root Reduction-to-all Alg…
HPC systems keep growing in size to meet the ever-increasing demand for performance and computational resources. Apart from increased performance, large scale systems face two challenges that hinder further growth: energy efficiency and…
Effective communication between the server and workers plays a key role in distributed optimization. In this paper, we focus on optimizing the server-to-worker communication, uncovering inefficiencies in prevalent downlink compression…
This article introduces a novel paradigm for the unsourced multiple-access communication problem. This divide-and-conquer approach leverages recent advances in compressive sensing and forward error correction to produce a computationally…
CP tensor decomposition with alternating least squares (ALS) is dominated in cost by the matricized-tensor times Khatri-Rao product (MTTKRP) kernel that is necessary to set up the quadratic optimization subproblems. State-of-art parallel…
Work introduces a hierarchical binary tree-based reduction that replaces standard self-attention. The core idea is to use a recursive Gated Linear Unit merge operation, achieving O(n) total merge operations O(log n) parallel depth O(n d^2)…
Optical circuit-switched networks have emerged as an appealing alternative to electrical fabrics as they can reconfigure the network topology at runtime, reducing communication cost and improving bandwidth utilization. Yet exploiting…
Given the increasing demand from wireless applications, designing energy-efficient group communication protocols is of great importance to multi-hop wireless networks. A group communication session involves a set of member nodes, each of…
We introduce a new theoretical framework for deriving lower bounds on data movement in bilinear algorithms. Bilinear algorithms are a general representation of fast algorithms for bilinear functions, which include computation of matrix…
It is a challenging task to train large DNN models on sophisticated GPU platforms with diversified interconnect capabilities. Recently, pipelined training has been proposed as an effective approach for improving device utilization. However,…
In this article, hybrid parallel bidirectional sieve method is implemented by SMP Cluster, the individual computational units joined together by the communication network, are usually shared-memory systems with one or more multicore…
Designing energy-efficient all-to-all multicasting protocols is of of great importance for multi-hop wireless networks such as wireless sensor networks and wireless ad hoc networks. In an all-to-all multicast session, there exists a set of…
Distributed Pseudo-tree Optimization Procedure (DPOP) is a well-known message passing algorithm that has been used to provide optimal solutions of Distributed Constraint Optimization Problems (DCOPs) -- a framework that is designed to…
In this paper, we address the scenario where nodes with sensor data are connected in a tree network, and every node wants to compute a given symmetric Boolean function of the sensor data. We first consider the problem of computing a…
In this paper, we present a new algorithm for parallel Monte Carlo tree search (MCTS). It is based on the pipeline pattern and allows flexible management of the control flow of the operations in parallel MCTS. The pipeline pattern provides…
This paper proposes a mobile pipeline computing concept in a Device-to-Device (D2D) communication setup and studies related issues, where D2D is likely based on millimeter-wave (mmWave) in the 5G mobile communication. The proposed…
Distributed processing frameworks, such as MapReduce, Hadoop, and Spark are popular systems for processing large amounts of data. The design of efficient algorithms in these frameworks is a challenging problem, as the systems both require…
In the Fully Sharded Data Parallel (FSDP) training pipeline, collective operations can be interleaved to maximize the communication/computation overlap. In this scenario, outstanding operations such as Allgather and Reduce-Scatter can…
We consider algorithms for going from a "full" matrix to a condensed "band bidiagonal" form using orthogonal transformations. We use the framework of "algorithms by tiles". Within this framework, we study: (i) the tiled bidiagonalization…
We present a new method for performing global redistributions of multidimensional arrays essential to parallel fast Fourier (or similar) transforms. Traditional methods use standard all-to-all collective communication of contiguous memory…
Building a spanning tree, minimum spanning tree (MST), and BFS tree in a distributed network are fundamental problems which are still not fully understood in terms of time and communication cost. x The first work to succeed in computing a…