Loading…
avatar for Ronan Keryell

Ronan Keryell

Xilinx
Principal Software Engineer
San José, California.
Post-modern C++-based programming models.
SYCL heterogeneous computing open standard from Khronos Group.
triSYCL implementation based on C++20, Clang & LLVM.
Code generation for FPGA & CGRA.
Wednesday, October 17
 

8:00am PDT

Registration & Breakfast 0 - Lobby Area

9:00am PDT

Welcome 1 - General Session (Rm LL20ABC)Tanya Lattner

9:15am PDT

The Future Direction of C++ and the Four Horsemen of Heterogeneous Computing 1 - General Session (Rm LL20ABC)Michael Wong

10:00am PDT

Break 0 - Lobby Area

10:30am PDT

Coroutine Representations and ABIs in LLVM 1 - General Session (Rm LL20ABC)John McCall

11:00am PDT

Stories from RV: The LLVM vectorization ecosystem 2 - Technical Talk (Rm LL21AB)Simon Moll

11:30am PDT

Outer Loop Vectorization in LLVM: Current Status and Future Plans 2 - Technical Talk (Rm LL21AB)Satish Guggilla • Florian Hahn

12:00pm PDT

Extending the SLP vectorizer to support variable vector widths 2 - Technical Talk (Rm LL21AB)Vasileios Porpodas

12:30pm PDT

Lunch 0 - Lobby Area

2:00pm PDT

Lightning Talks 1 - General Session (Rm LL20ABC)JF Bastien • Jeremy Bennett • Kirill Bobyrev • Kavon Farvardin • Mikhail R. Gadelha • Stefan Gränitz • Andrei Homescu • Jessica Paquette • Paul Robinson • Steve Scalpone • Steven Wu

2:30pm PDT

Round Tables 4 - Round Tables (Rm LL21EF)

3:00pm PDT

Lightning Talks 1 - General Session (Rm LL20ABC)Louis Dionne • Aleksandr Efremov • Hal Finkel • Erich Focht • James Henderson • Julie Hockett • Joshua Magee • Konstantina Mitropoulou • Bharathi Seshadri • Malhar Thakkar • Nico Weber

4:00pm PDT

Break 0 - Lobby Area

4:30pm PDT

Round Tables 4 - Round Tables (Rm LL21EF) Faster, Stronger C++ Analysis with the Clang Static Analyzer 1 - General Session (Rm LL20ABC)Artem Dergachev • George Karpenkov

5:00pm PDT

Memory Tagging, how it improves C++ memory safety, and what does it mean for compiler optimizations 1 - General Session (Rm LL20ABC)Kostya Serebryany
 
Thursday, October 18
 

8:15am PDT

Registration & Breakfast 0 - Lobby Area

9:15am PDT

Glow: LLVM-based machine learning compiler 1 - General Session (Rm LL20ABC)Roman Levenstein • Jakob Olesen • Nadav Rotem

10:00am PDT

Graph Program Extraction and Device Partitioning in Swift for TensorFlow 1 - General Session (Rm LL20ABC)Mingsheng Hong • Chris Lattner

10:30am PDT

Break 0 - Lobby Area

11:00am PDT

Art Class for Dragons: Supporting GPU compilation without metadata hacks! 2 - Technical Talk (Rm LL21AB)Neil Hickey

11:30am PDT

Profile Guided Function Layout in LLVM and LLD 1 - General Session (Rm LL20ABC)Michael Spencer How to use LLVM to optimize your parallel programs 2 - Technical Talk (Rm LL21AB)William Moses

12:00pm PDT

Optimizing Indirections, using abstractions without remorse. 1 - General Session (Rm LL20ABC)Johannes Doerfert

2:00pm PDT

Loop Transformations in LLVM: The Good, the Bad, and the Ugly 2 - Technical Talk (Rm LL21AB)Michael Kruse

2:30pm PDT

Implementing an OpenCL compiler for CPU in LLVM 1 - General Session (Rm LL20ABC)Evgeniy Tyurin

3:00pm PDT

Revisiting Loop Fusion, and its place in the loop transformation framework. 2 - Technical Talk (Rm LL21AB)Kit Barton

3:30pm PDT

Poster Session 0 - Lobby Area

4:30pm PDT

Ideal versus Reality: Optimal Parallelism and Offloading Support in LLVM 3 - BoF (Rm LL21CD)Vikram Adve • Johannes Doerfert • Hal Finkel • TB Schardl • Xinmin Tian

5:00pm PDT

Should we go beyond `#pragma omp declare simd`? 3 - BoF (Rm LL21CD)Francesco Petrogalli

5:30pm PDT

Closing 1 - General Session (Rm LL20ABC)Tanya Lattner