The Fifth Workshop on the LLVM Compiler Infrastructure in HPC (original) (raw)

Agenda

Room: D175

Time	Speaker	Title
9:00	Hal Finkel	Welcome
9:15	Bert Maher	Keynote: Glow: An Optimizing Compiler for High-Performance Machine Learning
10:00	Coffee	Break
10:30	Guray Ozen	OpenMP GPU Offload in Flang and LLVM
11:00	David Truby	Pointers Inside Lambda Closure Objects in OpenMP Target Offload Regions
11:30	Joel Denny	Clacc: Translating OpenACC to OpenMP in Clang
12:00	Francesco Petrogalli	LLVM and the automatic vectorization of loops invoking math routines: -fsimdmath
12:30	N/A	Lunch
2:00	Matt Masten	Function/Kernel Vectorization via Loop Vectorizer
2:30	Michael Kruse	User-Directed Loop-Transformations in Clang
3:00	Coffee	Break
3:30	Gábor Dániel Balogh	OP2-Clang: A Source-to-Source Translator Using Clang/LLVM LibTooling
4:00	Fabian Schlebusch	PInT: Pattern Instrumentation Tool for Analyzing and Classifying HPC Applications
4:30	Beau Johnston	AIWC: OpenCL-based Architecture Independent Workload Characterization
Lightning Talks:
5:00	Dorit Nuzman	Compiler Optimization for Heterogeneous Locality and Homogeneous Parallelism in OpenCL and LLVM
5:07	Lechen Yu	A Study of OpenMP Device Offloading in LLVM: Correctness and Consistency
5:14	Victor Lomuller	Challenges of C++ Heterogeneous Programming using SYCL Implementation Experience: the Four Horsemen of the Apocalypse
5:21	Final	Discussion
5:30	The	End

Sunday, November 11th, 6pm - 9pm: LLVM Social, and Flang Meetup Note: This is the evening before the workshop!Aloft Dallas Downtown 1033 Young St, Dallas, TX 75202Please RSVP to gklimowicz@nvidia.com (so that we can get a rough idea of headcount).
Tuesday, November 13th, 12:15pm - 1:15pm: BoF: LLVM in HPC: What’s New?Location: D171

Abstract

LLVM, winner of the 2012 ACM Software System Award, has become an integral part of the software-development ecosystem for optimizing compilers, dynamic-language execution engines, source-code analysis and transformation tools, debuggers and linkers, and a whole host of programming-language and toolchain-related components. Now heavily used in both academia and industry, where it allows for rapid development of production-quality tools, LLVM is increasingly used in work targeted at high-performance computing. Research in, and implementation of, program analysis, compilation, execution, and profiling has clearly benefited from the availability of a high-quality, freely-available infrastructure on which to build. This workshop will focus on recent developments, from both academia and industry, that build on LLVM to advance the state of the art in high-performance computing.

In cooperation with:

Held in conjunction with SC18: The International Conference for High Performance Computing, Networking, Storage, and Analysis

Format

This workshop will feature contributed papers and invited talks focusing on recent developments, from both academia and industry, that build on LLVM to advance the state of the art in high-performance computing.

Topics of interest include, but are not limited to:

Compiler design for highly-concurrent/parallel environments
Compilation techniques targeted at high-performance computing codes
Programming-language implementation techniques enabling high performance and high productivity
Embedding compilation and dynamic execution at scale
Tools for optimization, profiling, and feedback
Source code transformation and analysis
Gap analyses of open-source LLVM-based tools

The workshop will hold a lightning-talk session. Please contribute to making this session both vibrant and informative! An abstract and one-page summary are required for consideration.

Deadlines

Paper submissions due: ~~September 1, 2018~~ Extended to: September 9, 2018 (AoE)
Notification to authors of acceptance: September 28, 2018
Camera-ready papers due: October 8, 2018
Workshop takes place: November 12, 2018

Please see the SC18 home page for registration deadlines and other information associated with the parent event.

Submissions

Please submit papers using the SC18 Submissions system by selecting the "SC18 Workshop: LLVM-HPC2018 Full Papers" form. Papers must be in IEEE conference format (templates are available). Papers should be no more than 12 pages (including references and figures) and must be at least eight pages long. Please also note IEEE's Article-Posting Policy.

To submit a lightning talk, please use the "SC18 Workshop: LLVM-HPC2018 Lightning Talks" form.

Proceedings

The proceedings will be archived in IEEE Xplore through TCHPC. Lightning-talk summaries will not be included in the proceedings.

Organizers

Hal Finkel, Argonne National Laboratory

Program Committee

Name	Affiliation
Alexis Perry	Los Alamos National Laboratory
Cameron McInally	Cray
Chandler Carruth	Google
Erik Schnetter	Perimeter Institute for Theoretical Physics
Frank Winter	Thomas Jefferson National Accelerator Facility
James Brodman	Intel
Jeff Hammond	Intel
Jim Cownie	Intel
Keno Fischer	Julia Computing, Inc.
Michael Wong	Codeplay
Nadav Rotem	Facebook
Pat McCormick	Los Alamos National Laboratory
Ralf Karrenberg	NVIDIA
Sameer Shende	University of Oregon
Sunita Chandrasekaran	University of Delaware
Teresa Johnson	Google
Tobias Grosser	ETH Zürich
Torsten Hoefler	ETH Zürich

Contact Information

Hal Finkel (hfinkel@anl.gov)