Write a Blog >>
Events (13 results)

A Supernodal All-Pairs Shortest Path Algorithm

Main Conference When: Tue 25 Feb 2020 14:25 - 14:50 People: piyush kumar sao, Ramki Kannan, Prasun Gera, Rich Vuduc

… We show how to exploit graph sparsity in the \FW algorithm for the all-pairs shortest path (\APSP) problem. \FW is an attractive choice for \APSP on high-performing systems due to its structural similarity to solving dense linear …

spECK: Accelerating GPU Sparse Matrix-Matrix Multiplication Through Lightweight Analysis

Main Conference When: Wed 26 Feb 2020 11:20 - 11:45 People: Mathias Parger, Martin Winter, Daniel Mlakar, Markus Steinberger

… performance for certain types of matrices, but fail to accelerate all kinds … characteristics shows that we outperform all existing solutions in 85% of the cases …

Poster: A tool for top-down performance analysis of GPU-accelerated applications

Brief Announcements People: Keren Zhou, Mark W. Krentel, John Mellor-Crummey

… of both CPU and GPU activity to derive and attribute metrics at all levels …

Brief Announcement: Nonblocking Persistent Software Transactional Memory

Brief Announcements People: H. Alan Beadle, Wentao Cai, Haosen Wen, Michael Scott

… of memory are consistent at all times. This can be challenging in multithreaded …

Parallel Race Detection with Futures

Main Conference When: Tue 25 Feb 2020 12:10 - 12:35 People: Yifan Xu, Kyle Singer, I-Ting Angelina Lee

… The use of futures can generate arbitrary dependences in the computation, making it difficult to race detect efficiently. Algorithms proposed by prior work to race detect programs with futures all have to execute the program …

Poster: Identifying Scalability Bottlenecks for Large-Scale Parallel Programs with Graph Analysis

Brief Announcements People: Yuyang Jin, Haojie Wang, Xiongchao Tang, Torsten Hoefler, Xu Liu, Jidong Zhai

… dependencies needed for root-cause analyses. Tracing collects all information …

Scalable Top-K Retrieval with Sparta

Main Conference When: Mon 24 Feb 2020 11:45 - 12:10 People: Gali Sheffi, Dmitry Basin, Edward Bortnikov, David Carmel, Idit Keidar

… score is the sum of its scores for all query terms. Top-k retrieval is often used …

Oak: A Scalable Off-Heap Allocated Key-Value Map

Main Conference When: Mon 24 Feb 2020 10:00 - 10:25 People: Hagar Meir, Edward Bortnikov, Anastasia Braginsky, Dmitry Basin, Yonatan Gottesman, Eshcar Hillel, Idit Keidar, Eran Meir, Gali Sheffi

… ConcurrentNavigableMap). Oak allows concurrency among all map operations …

Universal Wait-Free Memory Reclamation

Main Conference When: Mon 24 Feb 2020 15:15 - 15:40 People: Ruslan Nikolaev, Binoy Ravindran

… to guarantee wait-freedom. As this extension is non-trivial, we discuss all

Brief Announcement: Memory-Friendly Lock-Free Bounded Queues

Brief Announcements People: Nikita Koval, Vitaly Aksenov

… -free bounded queue algorithm which requires two fundamental assumptions: all

Taming Unbalanced Training Workloads in Deep Learning with Partial Collective Operations

Main Conference When: Mon 24 Feb 2020 11:20 - 11:45 People: Shigang Li, Tal Ben-Nun, Salvatore Di Girolamo, Dan Alistarh, Torsten Hoefler

… half of the participants must contribute gradients before continuing, all

YewPar: Skeletons for Exact Combinatorial Search

Main Conference When: Tue 25 Feb 2020 16:10 - 16:35 People: Blair Archibald, Patrick Maier, Rob Stewart, Phil Trinder

… a systematic performance analysis of all 12 YewPar skeletons on standard instances of 7 …

Tutorial: Abstractions and Algorithms for Efficiently Programming NVRAMs

Workshops and Tutorials When: Sat 22 Feb 2020 08:00 - 12:00Sat 22 Feb 2020 13:00 - 17:00 People: Naama Ben-David, Guy Blelloch, Laxman Dhulipala, Michal Friedman, Yuanhao Wei, Yan Gu, Charles McGuffey

… . Experimentally, we run all of the algorithms on the largest publicly-available … that underlies all of the previously discussed models is affected by read-write …