Welcome to the IWOCL and SYCLcon 2021 program of events
Virtual Event Logistics
- Instructions for joining the live sessions can be found in your registration confirmation.
- If you are missing registration emails please see: What to check when you don’t receive Eventbrite emails
- A range of Slack channels are in use to facilitate discussions. The link to join these channels can be found in your original registration confirmation.
- All our authors have been asked to check their Slack channel from Wednesday 28th to answer any questions.
- Registration is now closed. The Slack channels will remain open to anyone who registered before the 29th April.
Proceedings
This year’s proceedings are available on the
ACM Digital Library
OneAPI Dev Summit
The First 2021 oneAPI Developer Summit
In this one-day virtual summit, you will hear speakers from industry and academia share their experience working on innovative cross-platform, multi-vendor architecture solutions developed on oneAPI.
Date: Monday 26 April
Start time: 9:00am BST, 10:00 CET | Duration: All day
Register: Please visit oneAPI Developer Summit to register and for additional information on the program and speakers.
Tutorials
Registered delegates will be sent instructions for joining these live tutorials and the associated Slack channel.
A Hands-On Introduction To SYCL
Tutorial Lead: Rod Burns, Codeplay
Presenters: Igor Vorobtsov, Intel. Aksel Alpay, University of Heidelberg. Ronan Keryell, Xilinx. Peter Zuzek, Codeplay and Gordon Brown, Codeplay
Date: Tuesday 27 April
Start time: 09:00 BST, 10:00 CET | Duration: Approx 7 Hours
discuss on slack: #discuss_tutorial_intro_to_sycl
SYCL | Paper ID: 113
Layers for OpenCL
Brice Videau, Argonne National Lab
Date: Tuesday 27 April
Start time: 15:00 BST, 16:00 CET, 07:00 PST | Duration: 2 hours
slides | tutorial on github | video presentation | discuss on slack: #discuss_tutorial_layers_for_opencl
OpenCL | Paper ID: 103
Live Panel Discussions
Registered delegates will be sent instructions for joining these live panel discussions and the associated Slack channel.
OpenCL Panel Discussion
Panel Chair: Simon McIntosh-Smith, University of Bristol
Date: Wednesday 28 April
Start time: 16:00 BST, 17:00 CET, 08:00 PST | Duration: 60 mins
on-demand video | discuss on slack: #discuss_panel_opencl
- Neil Trevett , Khronos and NVIDIA
- Jesse Natalie, Microsoft
- Paul Miller. Boris FX
- Ben Ashbourgh, Intel
- Kevin Petit, Arm
- Alastair Murray, Codeplay
OpenCL | Paper ID: P1 |
SYCL Panel Discussion
Panel Chair: Simon McIntosh-Smith, University of Bristol
Date: Thursday 29 April
Start time: 16:00 BST, 17:00 CET, 08:00 PST | Duration: 60 mins
on-demand video | discuss on slack: #discuss_panel_sycl
- Michael Wong, Codeplay
- Ronan Keryell, Xilinx
- Aksel Alpay, University of Heidelberg
- James Reinders, Intel
- Tom Deakin, University of Bristol
- Hal Finkel, U.S. Department of Energy
- Peter Thoman, University of Innsbruck
SYCL| Paper ID: P2 |
Khronos Advisory Panel Meetings – OpenCL and SYCL
These meetings are for members of the OpenCL and SYCL advisory panels. Existing members will be sent call-in instructions directly by the Khronos Group. See below for details on becoming an Advisor.
Advisory Panel Meeting – OpenCL
Date: Monday 26 April
Start time: 16:00 BST, 17:00 CET, 08:00 PST
Members only
Call-in instructions will be sent to members by the Khronos Group
SYCL Advisory Panel Meeting
Date: Tuesday 27 April
Start time: 16:00 BST, 17:00 CET, 08:00 PST
Members only
Call-in instructions will be sent to members by the Khronos Group
An Invitation to Become an OpenCL or SYCL Advisor
Khronos understands that to best meet the needs of the industry, standards such as OpenCL and SYCL should incorporate the requirements and feedback from industry experts. Consequently Khronos has established Advisory Panels to serve as a forum where said experts can engage in a bi-directional discourse with the corresponding Working Group (WG) to communicate their requirements, provide feedback on draft specs, consume WG materials offline, help prioritize work on features and among others, participate in shaping the ecosystem.
Research Papers and Technical Talks
Registered delegates will be able to view the on-demand video presentations of these sessions from Wednesday 28th April. All our speakers will be asked to check their Slack channel regularly to answer any questions coming in from the community.
KEYNOTE
SYCL, DPC++, XPUs, oneAPI – a View from Intel
Speaker: James Reinders, Intel.
on-demand video | slides | Discuss on slack: #discuss_talks_sycl | SYCL | Paper ID: K01 |
INVITED TALK
An Insight Into Kalray’s OpenCL Optimum Implementation
Speaker: Sebastien Leduc, Kalray. Software Engineering Director
on-demand video | slides | discuss on slack: #discuss_talks_opencl | OpenCL | Paper ID: K02 |
INVITED TALK
OpenCL Working Group – A State of the Union
Speaker: Neil Trevett, NVIDIA | Khronos Group, President and OpenCL Working Group Chair.
on-demand video | slides (updated) | discuss on slack: #discuss_talks_opencl | OpenCL | Paper ID: K03 |
INVITED TALK
SYCL Working Group – A State of the Union
Speaker: Michael Wong , Codeplay | Khronos Group, SYCL Working Group Chair.
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Paper ID: K04 |
BEST PAPER*
Sylkan: Towards a Vulkan Compute Target Platform for SYCL
Speaker: Peter Thoman, University of Innsbruck (UIBK)
Co-Authors: Daniel Gogl (UIBK); Thomas Fahringer (UIBK)
* The best paper award for 2021 was sponsored by Arm
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Paper ID: 14 |
Performance-Portable Distributed k-Nearest Neighbors using Locality-Sensitive Hashing and SYCL
Speaker: Marcel Breyer, University of Stuttgart, IPVS
Co-Authors: Gregor Daiß (University of Stuttgart, IPVS); Dirk Pflüger (University of Stuttgart, IPVS)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Paper ID: 24 |
Toward Performance Portability of Highly Parametrizable TRSM Algorithm Using SYCL
Speaker: Thales Sabino, Codeplay Software
Co-Authors: Mehdi Goli (Codeplay software Ltd)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Paper ID: 26 |
On Measuring the Maturity of SYCL Implementations by Tracking Historical Performance Improvements
Speaker: Wei-Chen Lin, University of Bristol
Co-Authors: Tom Deakin (University of Bristol); Simon McIntosh-Smith (University of Bristol)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Paper ID: 33 |
Experiences Supporting DPC++ in AMReX
Speaker: Sravani Konda, Intel Corporation
Co-Authors: Dunni Aribuki (Intel Corporation); Weiqun Zhang (Lawrence Berkeley National Laboratory); Kevin Gott (National Energy Research Scientific Computing Center); Christopher Lishka (Intel Corporation)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Tech Talk ID: 04 |
Developing Medical Imaging Application Across GPU, FPGA and, CPU Using oneAPI
Speaker: Wang Yong, Intel
Co-Authors: Zhou Yongfa (Intel); Wang Scott (Intel); Yang Wang (Intel Corporation); Xu Qing (Intel); Wang Chen (Intel)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Tech Talk ID: 11 |
hipSYCL in 2021: Peculiarities, Unique Features and SYCL 2020
Speaker: Aksel Alpay, Heidelberg University
Co-Authors: Vincent Heuveline (Heidelberg University)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Tech Talk ID: 23 |
Can SYCL and OpenCL Meet the Challenges of Functional Safety?
Speaker: Illya Rudkin, Codeplay Software
Co-Authors: Rod Burns (Codeplay Software)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Paper ID: 20 |
Experiences with Adding SYCL Support to GROMACS
Speaker: Andrey Alekseenko, KTH Royal Institute of Technology
Co-Authors: Szilárd Páll (KTH Royal Institute of Technology); Erik Lindahl (KTH Royal Institute of Technology)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Tech Talk ID: 22 |
Extending DPC++ with Support for Huawei Ascend AI Chipset
Speaker: Rasool Maghareh, Huawei Heterogeneous Compiler Lab
Co-Authors: Wilson Feng (Huawei Heterogeneous Compiler Lab); Kai-Ting Amy Wang (Huawei Heterogeneous Compiler Lab)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Tech Talk ID: 15 |
Toward a Better Defined SYCL Memory Consistency Model
Speaker: Ben Ashbaugh, Intel Corporation
Co-Authors: James C Brodman (Intel Corporation); Michael Kinsner (Intel Corporation); Gregory Lueck (Intel Corporation); John Pennycook (Intel Corporation); Roland Schulz (Intel Corporation)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | SYCL | Tech Talk ID: 28 |
Enabling OpenCL and SYCL for RISC-V Processors
Speaker: Colin Davidson, Codeplay Software
Co-Authors: Rod Burns (Codeplay Software); Aidan Dodds (Codeplay Software)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | OPENCL + SYCL | Paper ID: 19 |
Profiling Heterogeneous Computing Performance with VTune Profiler
Speaker: Vladimir Tsymbal, Intel
Co-Authors: Alexandr Kurylev (Intel)
on-demand video | slides | discuss on slack: #discuss_talks_sycl | OPENCL + SYCL | Tech Talk ID: 09 |
FAST: A Framework for High-Performance Medical Image Computing and Visualization
Speaker: Erik Smistad, Norwegian University of Science and Technology and SINTEF
on-demand video | slides | discuss on slack: #discuss_talks_opencl | OpenCL | Tech Talk ID: 17 |
Experiences Porting the SU3_Bench Microbenchmark to the Intel Arria 10 and Xilinx Alveo U280 FPGAs
Speaker: Douglas Doerfler, Lawrence Berkeley National Laboratory
on-demand video | slides | discuss on slack: #discuss_talks_opencl | OpenCL | Paper ID: 02 |
Accelerating Regular-Expression Matching on FPGAs with High-Level Synthesis
Speaker: Devon Callanan, University of Pittsburgh
on-demand video | slides | discuss on slack: #discuss_talks_opencl | OpenCL | Paper ID: 08 |
Performance Evaluation and Improvements of the PoCL Open-Source OpenCL Implementation on Intel CPUs
Speaker: Tobias Baumann, Zuse Institute Berlin
on-demand video | slides | discuss on slack: #discuss_talks_opencl | OpenCL | Paper ID: 30 |
Towards Evaluating High-Level Synthesis Portability and Performance Between Intel and Xilinx FPGAs
Speaker: Anthony M Cabrera, Oak Ridge National Laboratory
on-demand video | slides | discuss on slack: #discuss_talks_opencl | OpenCL | Paper ID: 31 |
Executing Graphs with OpenCL
Speaker: Erik Tomusk, Codeplay Software
on-demand video | slides | discuss on slack: #discuss_talks_opencl | OpenCL | Tech Talk ID: 12 |
Posters
Registered delegates will be able to view the on-demand video presentations of these posters sessions from Tuesday 27th April 2021. All our speakers will be asked to check their Slack channel regularly to answer any questions coming in from the community.
Enabling the Use of C++20 Unseq Execution Policy for OpenCL
Speaker: Po-Yao Chang, National Tsing Hua University
Co-Authors: Tai-Liang Chen (National Tsing Hua University); Jenq-Kuen Lee (National Tsing Hua University)
on-demand video | poster | discuss on slack: #discuss_posters | OpenCL | Poster ID: 05 |
Experimenting with C++ Libraries in OpenCL Kernel Code
Speaker: Ole M Strohm, Arm
Co-Authors: Anastasia Stulova, Arm
on-demand video | slides | discuss on slack: #discuss_posters | OpenCL | Poster ID: 06 |
Trip Down the Compute Pipeline
Speaker: Lukasz Towarek, Intel Corporation
on-demand video | slides | discuss on slack: #discuss_posters | OpenCL | Poster ID: 07 |
Machine Learning Training with Tensor Virtual Machine (TVM) and Adreno GPUs
Speaker: Siva Rama Krishna Reddy, Qualcomm
Co-Authors: Hongqiang Wang (Qualcomm); Adarsh Golikeri (Qualcomm); Alex Bourd (Qualcomm)
on-demand video | slides | discuss on slack: #discuss_posters | OpenCL | Poster ID: 34 |
SYCL for Vitis 2020.2: SYCL & C++20 on Xilinx FPGA
Speaker: Gauthier Harnisch, Xilinx
Co-Authors: Andrew Gozillon (University of the West of Scotland); Ronan Keryell, (Xilinx), Lin-Ya Yu (Xilinx, Inc); Ralph Wittig (Xilinx); Luc Forget (Xilinx)
poster | discuss on slack: #discuss_posters | SYCL | Poster ID: 27 |
Bringing SYCL to Ampere Architecture
Speaker: Steffen Larsen, Codeplay Software
Co-Authors: Rod Burns (Codeplay Software); Brandon Cook (Lawrence Berkeley National Laboratory); Douglas Doerfler (Lawrence Berkeley National Laboratory); Kevin Harms (Lawrence Berkeley National Laboratory); Thomas Applencourt (Lawrence Berkeley National Laboratory); Stuart Adams (Codeplay Software)
on-demand video | slides | discuss on slack: #discuss_posters | SYCL | PosterID: 16 |
Path Tracing on FPGA with SYCL and C++20
Speaker: Luc Forget (INRIA)
Co-Authors: Krishna Kumar Ranipet Murugan (NCSU); Ronan Keryell, (Xilinx); Gauthier Harnisch (Xilinx);
poster | discuss on slack: #discuss_posters | SYCL | Poster ID: 25 |
SYCL for Xilinx Versal ACAP AIE CGRA
Speaker: Ronan Keryell, Xilinx
Co-Authors: Andrew Gozillon (University of the West of Scotland); Gauthier Harnisch (Xilinx); Hyun Kwon (Xilinx); Ravikumar Chakaravarthy (Xilinx); Ralph Wittig (Xilinx)
poster | discuss on slack: #discuss_posters | SYCL | Poster ID: 29 |
Approaching Coupled Cluster Theory with Perturbative Triples using SYCL
Speaker: Abhishek Bagusetty, Argonne National Laboratory
Co-Authors: Jinsung Kim (Pacific Northwest National Laboratory); Ajay Panyala (Pacific Northwest National Laboratory); Alvaro Vazquez-Mayagoitia (Argonne National Laboratory); Karol Kowalski (Pacific Northwest National Laboratory); Sriram Krishnamoorthy (Pacific Northwest National Laboratory)