2023-06-05 OneAPI Hybrid Workshop (hoap1s23)

Course	OneAPI Hybrid Workshop
Number	hoap1s23
Available places	32
Date	05.06.2023 – 07.06.2023
Price	€ 0.00
Location	HYBRID: ONLINE/LRZ Boltzmannstr. 1 85748 Garching near Munich
Room	Hörsaal
Registration deadline	26.05.2023 23:55
E-mail	education@lrz.de

This course will be organised as a hybrid event with the possibility to attend at LRZ in Garching near Munich or online. On-site attendence will allow for participation in the exercises in the afternoon and for direct interaction with trainers during hands-on and demos. If you prefer to attend online you cannot participate in the exercises. The event will only be streamed in the morning.

There will be on-site catering service with lunch for free. We also plan to visit a restaurant/ beergarden on the first evening. You have the opportunity to attend a guided tour of the LRZ Computing Cube on the second afternoon. When registering, you can choose between 2 booking options:

- Participants ONLINE: If you want to attend remotely. You cannot participate in the exercises.
- Participants ON-SITE: If you want to attend at LRZ in Garching near Munich.

On-site participants are expected to bring their own laptops.

We welcome you to join us in learning about the oneAPI initiative and how to get on a standards-based path for mixed architectures programming. This 3-day jointly prepared LRZ / Intel hybrid (combined online and onsite) workshop will teach the participants about the latest and expanded features of the Intel oneAPI toolkits.
The presenters will show you how to write applications on heterogenous hardware platforms.

Who should attend?

C/C++/ Fortran application developers
GPU, CPU & HPC programmers

Why should you attend this workshop?

Develop best-practice heterogenous programming techniques from hands-on tutorials and tech talks delivered by tools specialists on oneAPI software development tools
Learn how you can efficiently use oneAPI Toolkits on various platforms such as Intel, AMD and NVIDIA for HPC and enterprise workloads
Understand how to use the C++ based SYCL Compiler to develop heterogenous applications running on CPUs and HW accelerators from multiple vendors
Learn to migrate CUDA codes to SYCL
How to squeeze most performance out of your application using the oneAPI optimization tools
Benefit from the Intel Dev Cloud development platform/sandbox

Topics covered:

Intel oneAPI – Introduction to Heterogenous Development Environment
Introduction to the Intel Dev Cloud for Testing and Porting Applications
Direct programming with oneAPI SYCL ++ and Fortran/OpenMP (with demos)
OneAPI Compatibility Tool for porting CUDA applications to DPC++ (with demos)
Intel oneAPI libraries oneMKL, oneDAL and others (with demos)
Intel OpenMP 5 for Offloading to Accelerators in Fortran codes (with demos)
Intel Hardware Roadmap/XPU and Architecture Specifics
Application Profiling for Heterogenous Hardware (with demos)

This Workshop is co-organised by the Leibniz Supercomputing Centre (LRZ) and Intel for EuroCC@GCS, the German National Competence Centre for High-Performance Computing.

Prerequisites

On-site participants are expected to bring their own laptops.
It is assumed that participants have a basic knowledge about HPC and should be able to program in C/C++ and/or Fortran.

Hands-On

There will be practical exercises after lunch.
Attendees will be given access to an Intel cluster.

Extras

On-site catering service with lunch for free.
A common visit of a restaurant/ beergarden on the first evening.
A guided tour of the LRZ Computing Cube on the second afternoon. The tour will take place on site at the LRZ building and will not be streamed into the internet.

Daily Agenda (Subject to Change) (all Times in CEST)

Day 1 - Monday
09:00 - 09:10	Welcome and Introduction to Day 1
09:10 - 09:30	OneAPI – Introduction to a mixed Architecture Development Environment - Motivation and oneAPI Standardization - Intel’s oneAPI Toolkits Portfolio and Components - Intel oneAPI plug-ins for Nvidia and AMD hardware (CPU and GPUs)
09:30 - 09:50	Introduction to the DevCloud - A sandbox for software development and benchmarking - Purpose: Demoing, testing and porting applications - Hardware and Software offerings - How to onboard & how to get an DevCloud account
09:50 - 10:40	Direct programming with oneAPI Compilers (Part 1) – with Demos - Intro to heterogenous programming model with SYCL 2020 - SYCL features and examples - “Hello World” Example - Device Selection - Execution Model - Compilation and Execution Flow - Memory Model; Buffers, Unified Shared Memory (USM) - Performance optimizations with SYCL features
10:40 - 11:10	Coffee break
11:10 - 12:00	Direct programming with oneAPI Compilers (Part 2) – with Demos - Intro to heterogenous programming model with SYCL 2020 - SYCL features and examples - “Hello World” Example - Device Selection - Execution Model - Compilation and Execution Flow - Memory Model; Buffers, Unified Shared Memory (USM) - Performance optimizations with SYCL features
12:00 - 12:15	oneAPI SYCL Case Study - SeiSsol
12:15 - 12:30	oneAPI SYCL Case Study - DPEcho
12:30 - 13:30	Lunch (Buffet Lunch in front of Auditorium)
13:30 - 14:45	Hands-on session applying SYCL and DPPC++ knowledge (heterogenous programming) Practical excercises on dedicated target systems: - Install oneAPI tools - Write your first SYCL programs - Develop own kernels - Test your own heterogenous programs - How to avoid errors
14:45 - 15:15	Coffee break
15:15 - 16:30	Hands-on session - Programming with SYCL (continued) Practical excercises on dedicated target systems
16:20 - 16:45	Questions and Answers - Wrap up of the day
17:30 - 21:00	Social Networking Event (Restaurant or Beergarden)
Day 2 - Tuesday
09:00 - 09:05	Welcome and Introduction to Day 2
09:05 - 10:05	Intel OpenMP for Offloading – with Demos Parallelizing heterogenous applications with OpenMP 5.1
10:05 - 10:40	Intel oneAPI libraries (oneMKL) for HPC - with Demos Performance optimized libraries for numerical simulations and other purposes
10:40 - 11:10	Coffee break
11:10 - 11:50	Target NVIDIA and AMD with oneAPI and SYCL Using SYCL based NVIDIA and AMD plugins with Demos
11:50 - 12:20	Open Source Compatibility tool for porting purposes(SYCLomatic) - with Demo Migration Cuda based GPU Applications to SYCL
12:20 - 12:45	Intel Debugging Tools for heterogenous programming (CPU, GPU) - with Demos
12:45 - 13:15	Programming for Distributed HPC Systems using Intel MPI
13:15 - 14:15	Lunch (Buffet Lunch in front of Auditorium)
14:15 - 15:30	Hands-on session - Open MP - Applying MKL lib with examples
15:30 - 16:00	Coffee break
16:00 - 16:45	Hands-on session Using DPC+ Compatibility Tool
16:45 - 16:55	Questions and Answers - Wrap up of the day
17:00 - 18:00	Guided tour of the LRZ Computing Cube
Day 3 - Wednesday
09:00 - 09:05	Welcome and Introduction to Day 3
09:05 - 10:05	Application profiling for CPU and or mixed hardware withe the Intel VTune - Demos - Vtune general / main functionality (Hot spot analysis, …) starting with CPU - Profiling Tools Interfaces for GPU - Profile heterogenous SYCL/OpenMP Workloads with Intel VTune Profiler
10:05 - 10:20	Coffee break
10:20 - 11:20	Application profiling for CPU and or mixed hardware withe the INtel VTune - Demos - Vtune general / main functionality (Hot spot analysis, …) starting with CPU - Profiling Tools Interfaces for GPU - Profile heterogenous SYCL/OpenMP Workloads with Intel VTune Profiler
11:20 - 11:35	Coffee break
11:35 - 12:35	Application profiling for CPU and mixed hardware with the Intel Advisor - Demos - Advisor's main functionality (Vectorization and Roofline) starting with CPU - Estimate performance potential gains with Offload Advisor (CPU -> HW Accelerator) - Analyse heterogenous SYCL/OpenMP Workloads with Intel Advisor and Roofline analysis
12:35 - 14:05	Lunch (Buffet Lunch in front of Auditorium)
14:05 - 15:05	Hands-on session VTune - CPU focus - Identifying hotspots and optimize apps - Using the Profiler for System optimisation
15:05 - 15:25	Coffee break
15:25 - 16:25	Hands-on session VTune (continued) - CPU focus
16:25 - 16:30	Questions and Answers - Wrap up of the day

Language

English

Lecturers

Edmund Preiss, Igor Vorobstov, Gennady Fedorov, Heinrich Bockhorst, Rafael Lago, Matthias Kirchhart, Alina Shadrina, Tobias Kloeffel, Dmitry Sivkov, Rod Burns, Stephen Blair-Chappell, Pascal Rene Bähr

Salvatore Cielo (LRZ), Ravil Dorozhinskii (TUM)

Prices and Eligibility

The course is open and free of charge for all participants from academia and industry.

Registration

Please register with your official e-mail address to prove your affiliation.

Withdrawal Policy

See Withdrawal

Legal Notices

For registration for LRZ courses and workshops we use the service edoobox from Etzensperger Informatik AG (www.edoobox.com). Etzensperger Informatik AG acts as processor and we have concluded a Data Processing Agreement with them.

See Legal Notices

No.	Date	Time	Leader	Location	Room	Description
1	05.06.2023	09:00 – 12:20	LRZ Events	HYBRID: ONLINE/LRZ	Hörsaal	OneAPI Hybrid Workshop Lectures onsite and online Day 1
2	05.06.2023	13:20 – 16:30	LRZ Events	Leibniz Rechenzentrum	Seminarraum 1	OneAPI Hybrid Workshop Hands-on only onsite Day 1
3	05.06.2023	17:15 – 21:00	LRZ Events	Leibniz Rechenzentrum		OneAPI Hybrid Workshop Social Networking Event (Restaurant or Beergarden) Day 1
4	06.06.2023	09:00 – 13:15	LRZ Events	HYBRID: ONLINE/LRZ	Hörsaal	OneAPI Hybrid Workshop Lectures onsite and online Day 2
5	06.06.2023	14:15 – 16:55	LRZ Events	Leibniz Rechenzentrum	Seminarraum 1	OneAPI Hybrid Workshop Hands-on only onsite Day 2
6	06.06.2023	17:00 – 18:00	LRZ Events	Leibniz Rechenzentrum		OneAPI Hybrid Workshop Guided tour of the LRZ Computing Cube Day 2
7	07.06.2023	09:00 – 12:35	LRZ Events	HYBRID: ONLINE/LRZ	Hörsaal	OneAPI Hybrid Workshop Lectures onsite and online Day 3
8	07.06.2023	14:05 – 16:30	LRZ Events	Leibniz Rechenzentrum	Seminarraum 1	OneAPI Hybrid Workshop Hands-on only onsite Day 3

Contents