HPC workshop on heterogeneous architectures with SYCL / INTEL ONEAPI

Europe/Paris
Ecole Polytechnique - Aile 0 - Ground Floor

Ecole Polytechnique - Aile 0 - Ground Floor

Conference Room "Jean Lascoux" Route de Saclay Palaiseau
Laurent SERIES, Yannick FITAMANT (CNRS)
Description

HPC workshop on heterogeneous architectures with SYCL / INTEL ONEAPI

 

 

 

 

Purpose of workshop

Take advantage of architectures with new programming paradigms

In recent years computation has been exposed to a widely growing heterogeneity of the types of used devices and their manufacturers. This trend is driven by the specialisation of devices to fit the needs of specific workloads, which in turn is the strategy to satisfy the ever-growing demand in compute power. Heterogeneity is present in both high-performance computing and consumer electronics. Today's systems use a multitude of co-processors and accelerators, such as GPUs, TPUs, and FPGAs, in addition to the traditional CPU.

However, there isn't a simple, portable and efficient method to develop for these systems. Intel oneAPI (SYCL implementation) aims to fill this role.

Getting the maximum achievable performance out of today’s hardware is a fine balance between optimal use of underlying hardware features and using code that is portable, easily maintainable, and power-efficient. These factors don't necessarily work in tandem. They require prioritizing based on user needs. It's non-trivial for users to maintain separate code bases for different architectures. A standard, simplified programming model that can run seamlessly on scalar, vector, matrix, and spatial architectures will give developers greater productivity through increased code reuse and reduced training investment.

Improve yours Artificial Intelligence workflows

For years Intel has been developing tools and approaches to help Data Scientists and AI practitioners solve their everyday problems and accelerate the standard workflows. Come and see how Intel hardware has evolved to include specific features to accelerate AI workloads, including classical machine learning and Deep Learning. Learn how to make your existing AI code run faster on the CPU by using Intel tools and techniques. Discover how to improve your workflows and security

People

Researchers and engineers with HPC knowledge

Duration 

  • 4 days with custom-made sessions training (you choose)
  • indicative timetable here

Venue

Ecole Polytechnique / Palaiseau 

Registration

  • Open for plenary session on morning, monday 28 (30 persons max.)
  • Mandatory for all others sessions (15 persons max / session)

 

Even if you have been offered to register for each half-day (session), it is preferable to register for the full day (morning and afternoon session) on Tuesday, Wednesday and Thursday.

 

Click here to proceed to registration.

Prerequisites

Knowledge : The workshop aims at participants who are familiar with the C/C++ or Fortran programming languages and have working experience with the Linux operating system and the use of the command line. Experience with parallel programming is required

 

Hardware : bring your laptop with vscode

 

Course language is French.

Speakers

Inscription
Registration
    • 09:45
      Workshop Welcome
    • Morning session: Plenary session
      • 1
        General public introduction

        Heterogeneous programming

      • 11:00
        Coffee break
    • 12:30
      Lunch
    • Afternoon session
      • 2
        Program the heterogeneous environment without constructor dependency with SYCL with examples of the code
      • 15:00
        Coffee break
    • Morning session
      • 3
        Overview of Intel oneAPI compute libraries and compatibility tools
        • Presentation of oneDPL, oneTBB, oneMKL compute libraries
        • Compatibility Tool (CUDA/SYCL conversion)
      • 11:00
        Coffee break
    • 12:00
      Lunch
    • Afternoon session
      • 4
        Tutorial on using the Compatibility Tool
      • 15:00
        Coffee break
    • Morning session
      • 5
        Profiling and analyzing code performance using Vtune/Advisor
      • 11:00
        Coffee break
    • 12:00
      Launch
    • Afternoon session
      • 6
        Practical work code optimization
      • 15:00
        Coffee break
    • Morning session
      • 7
        Overview of Intel Tools for Artificial Intelligence

        For years Intel has been developing tools and approaches to help Data Scientists and AI practitioners solve their everyday problems and accelerate the standard workflows. Come and see how Intel hardware has evolved to include specific features to accelerate AI workloads, including classical machine learning and Deep Learning. Learn how to make your existing AI code run faster on the CPU by using Intel tools and techniques. Discover how to improve your workflows and security.

      • 11:00
        Coffee break
    • 12:00
      Lunch
    • Afternoon session
      • 8
        Practil work AI
      • 15:00
        Coffee break