The LLVM Compiler Infrastructure
Site Map:
Download!
Search this Site


Useful Links
Release Emails
14.0.6: Jun 2022
14.0.5: Jun 2022
14.0.4: May 2022
14.0.0: Mar 2022
13.0.1: Feb 2022
13.0.0: Oct 2021
12.0.1: Jul 2021
12.0.0: Apr 2021
11.0.1: Jan 2021
All Announcements

Maintained by the
llvm-admin team
The Second Workshop on LLVM in Parallel Processing (LLPP) will be held at ICPP 2022

The LLVM framework is a vast ecosystem that stretches far beyond a "simple" C/C++ compiler. The variety of programming-language and toolchain-related parts includes the components to support a host of parallel programming models and libraries, including CUDA, HIP, OpenMP, SYCL, and the C++ parallel algorithms. In addition, LLVM serves as a vehicle for various languages in which parallelism is a first class citizen, such as Julia or Chapel. The workshop is held in conjunction with the ICPP during the workshop day. If you are interested in attending the workshop, please register at the ICPP website.


  1. Schedule
  2. Contact
  3. About
  • Conference Dates: August 29th - September 1st, 2022
  • Workshop Date: August 29th, 2022: 13:30 - 17:00 CEST
  • Location: virtual
  • Coordinates: https://argonne.zoomgov.com/j/1616646016
Schedule

Time (EDT)

Speaker

Title

8:00 - 8:05 (5 min)

Giorgis Georgakoudis
Johannes Doerfert

Welcome and introduction

8:05 - 8:30 (25 min)

Delaram Talaashrafi

A Pipeline Pattern Detection Technique in Polly

8:30 - 9:15 (45 min)

William Moses

Enzyme: Automatic Differentiation for Parallel Programs

9:15 - 9:35 (20 min)

Break

9:35 - 10:00 (25 min)

Guilherme Valarini

The OpenMP Cluster Programming Model

10:00 - 10:20 (20 min)

Rafael Herrera

Automatic Asynchronous Execution of Synchronously Offloaded OpenMP Target Regions

10:20 - 10:40 (20 min)

Jan Hueckelheim

ORAQL: Optimistic Responses to Alias Queries in LLVM

10:40 - 11:00 (20 min)

Alister Johnson

MARTINI: The Little Match and Replace Tool for Automatic Code Rewriting

Topics of Interest
Focus areas LLPP include but are not limited to:

  • Parallelism-aware compilation and optimization
    • Parallelism analysis in compilation
    • Parallelism-aware compiler optimizations
  • Parallel programming languages and models
    • Domain-specific parallel programming languages
    • Novel parallelism models and implementations
  • Performance analysis and tuning in LLVM
    • Performance analysis tools for parallelism
    • Tuning tools for parallel execution

Proceedings
Accepted papers will be published by the ACM International Conference Proceedings Series (ICPS), in conjunction with those of other ICPP workshops, in a volume entitled 51st International Conference on Parallel Processing Workshops (ICPP 2022 Workshops). This volume will be available for download via the ACM Digital Library.

Submission Guidelines
LLPP follows the same submission guidelines as the ICPP conference, please visit ICPP CFP. In brief, paper submissions should be in the US letter ACM format, not exceed 10 pages (incl. references), prepared for double-blind review, and submitted in the PDF document format.

Important Dates (AoE)
Paper submission: June 17th, 2022
Author notification: July 1st, 2022 July 4th, 2022
Camera-ready submission: July 11th, 2022
Workshop date: August 29th, 2022

Organizing Committee
Johannes Doerfert, Argonne National Laboratory, USA
Giorgis Georgakoudis, Lawrence Livermore National Laboratory, USA

Program Committee
Deepak Eachempati, HPE, USA
EJ Park, Qualcomm, USA
Florian Hahn, Apple, UK
Jan Huckelheim, Argonne National Laboratory, USA
Kiran Chandramohan, ARM, UK
Konstantinos Parasyris, Lawrence Livermore National Laboratory, USA
Marc Jasper, Lawrence Livermore National Laboratory, USA
William Moses, Massachusetts Institute of Technology, USA

Contact

In case of any queries please reach out to the workshop organizers: Johannes Doerfert (jdoerfert@anl.gov) and Giorgis Georgakoudis (georgakoudis1@llnl.gov).

About
The workshop will have proceedings. More information on the paper format is to follow. Submissions can be made on the LLPP submission page. We will consider invited talks, tutorials, panels, and other non-research-paper submissions. Please reach out if you are interested.