Výsledky vyhledávání - Theory of computation Models of computation Concurrency

Skywalker: Efficient Alias-Method-Based Graph Sampling and Random Walk on GPUs Autor Wang, Pengyu, Li, Chao, Wang, Jing, Wang, Taolei, Zhang, Lu, Leng, Jingwen, Chen, Quan, Guo, Minyi

“…Graph sampling and random walk operations, capturing the structural properties of graphs, are playing an important role today as we cannot directly adopt…”
Získat plný text

Managing GPU Concurrency in Heterogeneous Architectures Autor Kayiran, Onur, Nachiappan, Nachiappan Chidambaram, Jog, Adwait, Ausavarungnirun, Rachata, Kandemir, Mahmut T., Loh, Gabriel H., Mutlu, Onur, Das, Chita R.

“…), and discuss the limitations of existing GPU-based concurrency management techniques when employed in heterogeneous systems…”
Získat plný text

In search of lost time: Axiomatising parallel composition in process algebras Autor Aceto, Luca, Anastasiadi, Elli, Castiglioni, Valentina, Ingolfsdottir, Anna, Luttik, Bas

“…This survey reviews some of the most recent achievements in the saga of the axiomatisation of parallel composition, along with some classic results. We focus…”
Získat plný text

A distributed operational view of Reversible Prime Event Structures Autor Melgratti, Hernan, Mezzina, Claudio Antares, Michele Pinna, G.

“…Reversible prime event structures extend the well-known model of prime event structures to represent reversible computational processes…”
Získat plný text

On Thin Air Reads Towards an Event Structures Model of Relaxed Memory Autor Jeffrey, Alan, Riely, James

“…This is the first paper to propose a pure event structures model of relaxed memory…”
Získat plný text

Gluon-Async: A Bulk-Asynchronous System for Distributed and Heterogeneous Graph Analytics Autor Dathathri, Roshan, Gill, Gurbinder, Hoang, Loc, Jatala, Vishwesh, Pingali, Keshav, Nandivada, V. Krishna, Dang, Hoang-Vu, Snir, Marc

“…) programming and execution model. BSP permits bulk-communication and uses large messages which are supported efficiently by current message transport layers, but bulk-synchronization can exacerbate the performance impact of load…”
Získat plný text

Virtualizing Transactional Memory Autor Rajwar, Ravi, Herlihy, Maurice, Lai, Konrad

“… If the transactional model is to gain wide acceptance, hardware support for transactions must be virtualized to hide these limitations in much the same…”
Získat plný text

Quantitative Reasoning for Proving Lock-Freedom Autor Hoffmann, Jan, Marmar, Michael, Zhong Shao

“…This article describes a novel quantitative proof technique for the modular and local verification of lock-freedom. In contrast to proofs based on temporal…”
Získat plný text

Flexible Decoupled Transactional Memory Support Autor Shriraman, Arrvindh, Dwarkadas, Sandhya, Scott, Michael L.

“…A high-concurrency transactional memory (TM) implementation needs to track concurrent accesses, buffer speculative updates, and manage conflicts…”
Získat plný text

Alchemist: A Transparent Dependence Distance Profiling Infrastructure Autor Zhang, Xiangyu, Navabi, Armand, Jagannathan, Suresh

“… This paper addresses important aspects of this issue by proposing a novel profiling technique to automatically detect available concurrency in C programs…”
Získat plný text

Bandwidth Bandit: Quantitative characterization of memory contention Autor Black-Schaffer, David, Nikoleris, Nikos, Hagersten, Erik, Eklov, David

“…On multicore processors, co-executing applications compete for shared resources, such as cache capacity and memory bandwidth. This leads to suboptimal resource…”
Získat plný text

Decidability and Complexity for Quiescent Consistency Autor Dongol, Brijesh, Hierons, Robert M.

“…Quiescent consistency is a notion of correctness for a concurrent object that gives meaning to the object's behaviours in quiescent states, i.e., states in…”
Získat plný text

Network-Offloaded Bandwidth-Optimal Broadcast and Allgather for Distributed AI Autor Khalilov, Mikhail, Girolamo, Salvatore Di, Chrapek, Marcin, Nudelman, Rami, Bloch, Gil, Hoefler, Torsten

“…In the Fully Sharded Data Parallel (FSDP) training pipeline, collective operations can be interleaved to maximize the communication/computation overlap…”
Získat plný text

Fast Track: A Software System for Speculative Program Optimization Autor Kelsey, Kirk, Bai, Tongxin, Ding, Chen, Zhang, Chengliang

“…Fast track is a software speculation system that enables unsafe optimization of sequential code. It speculatively runs optimized code to improve performance…”
Získat plný text

Asynchronous Distributed-Memory Parallel Algorithms for Influence Maximization Autor Singhal, Shubhendra Pal, Hati, Souvadra, Young, Jeffrey, Sarkar, Vivek, Hayashi, Akihiro, Vuduc, Richard

“…Influence maximization (IM) is the problem of finding the k most influential nodes in a graph. We propose distributed-memory parallel algorithms for the two…”
Získat plný text

MDLoader: A Hybrid Model-Driven Data Loader for Distributed Graph Neural Network Training Autor Bae, Jonghyun, Choi, Jong Youl, Pasini, Massimiliano Lupo, Mehta, Kshitij, Zhang, Pei, Ibrahim, Khaled Z.

“… MDLoader features a model-driven performance estimator that dynamically selects between one-sided and collective communication at the beginning…”
Získat plný text

A static verification framework for message passing in Go using behavioural types Autor Lange, Julien, Ng, Nicholas, Toninho, Bernardo, Yoshida, Nobuko

“… This work proposes a practical verification framework for message passing concurrency in Go by developing a robust static analysis that infers an abstract model of a program's communication behaviour…”
Získat plný text

NEO-DNND: Communication-Optimized Distributed Nearest Neighbor Graph Construction Autor Iwabuchi, Keita, Steil, Trevor, Priest, Benjamin W., Pearce, Roger, Sanders, Geoffrey

“…Graph-based approximate nearest neighbor algorithms have shown high neighbor structure representation quality. NN-Descent is a widely known graph-based…”
Získat plný text

Asynchronous Template Games and the Gray Tensor Product of 2-Categories Autor Mellies, Paul-Andre

“…, trajectories and reshufflings (or reschedulings) as 2-cells. In order to achieve the purpose, we take seriously the parallel between asynchrony in concurrency and the Gray tensor product of 2-categories…”
Získat plný text

Parallel random numbers: as easy as 1, 2, 3 Autor Salmon, John K., Moraes, Mark A., Dror, Ron O., Shaw, David E.

“…Most pseudorandom number generators (PRNGs) scale poorly to massively parallel high-performance computation because they are designed as sequentially dependent state transformations…”
Získat plný text