HPC SAUDI 17 Conference Tutorials will provide eight different sessions to the attendees at KAUST. These half- or full-day tutorials are suitable for new, intermediate and advanced users, managers, students, systems administrators and anyone seeking an overview of parallel computing; debugging, performance and optimization tools; scientific applications; use of compute accelerators (Xeon Phi, GPUs), systems administration and resource management; procuring HPC systems.
Please click here to register to the tutorials you plan to attend. A detailed agenda can be found here.
Here is a brief description of the different sessions:
Morning sessions:
• KSL: HPC 101: How to use a Supercomputer?
Goal: Provide new users with general knowledge on accessing Shaheen, including compilers, libraries, software available to users and using examples to schedule and run efficiently the desired application. Several examples using chemistry, CFD, climate and biology packages along with visualization tools will be presented.
Duration: Half-day
Level: Beginner
• Cray: Performance tuning and scaling on Shaheen
Goal: Help users to run their applications efficiently and effectively at scale. The goal is to tune the performance of such codes by detecting, locating and eventually solving performance bottlenecks with a focus on communication and I/O.
Duration: Half-day
Level: Intermediate to advanced
• Intel: Vectorization and performance characterization using Intel Vector Advisor
Goal: Review the basics of vectorization for Intel processors through examples by detecting the bottlenecks, pinpointing vectorization problems, to fixing the code, applying OpenMP pragmas, changing memory layout and other techniques basing on Intel Advisor tools data and recommendations.
Duration: Full-day
Level: Intermediate to advanced
• Containers: Portable, repeatable user-oriented application delivery. Build, ship, run any app anywhere
Goal: Provide an overview on the containers platform, including system components: images, containers, repositories, clustering and orchestration and demonstrate through live demo and hands-on exercises, the reuse case of containers in building a portable distributed application cluster running a variety of workloads, including HPC workload.
Duration: Half-day
Level: Intermediate to advanced
• SLURM: Intermediate Slurm Training for System Administrators
Goal: Cover installing and configuring the Slurm workload manager on a Linux cluster, including a review of Slurm’s internal architecture, discussion of different capabilities and features and a live demo of configuring Slurm for various workload.
Duration: Full-day
Level: Intermediate to advanced
Afternoon sessions:
• Allinea: Efficient HPC Development and Production with Allinea Tools
Goal: Help scientific users from the development of an application to the production by debugging, profiling your applications with Allinea tools.
Duration: Half-day
Level: Intermediate to advanced
• Intel: Vectorization and performance characterization using Intel Vector Advisor
Goal: Review the basics of vectorization for Intel processors through examples by detecting the bottlenecks, pinpointing vectorization problems, to fixing the code, applying OpenMP pragmas, changing memory layout and other techniques basing on Intel Advisor tools data and recommendations.
Duration: Full-day
Level: Intermediate to advanced
• Nvidia: How to get started with Deep Learning
Goal: Learn how you can apply deep learning in image recognition and segmentation for developers and data scientists worldwide how to solve challenging problems with deep learning.
Duration: Half-day
Level: Beginner to Intermediate
• SLURM: Intermediate Slurm Training for System Administrators
Goal: Cover installing and configuring the Slurm workload manager on a Linux cluster, including a review of Slurm’s internal architecture, discussion of different capabilities and features, and live demo of configuring Slurm for various workload.
Duration: Full-day
Level: Intermediate to advanced
• RedOak: Business aspects of HPC: acceptance, installation & commissioning
Goal: Guide attendees through the process of purchasing and deploying an HPC system. It will cover the whole process from engaging with stakeholders in securing funding, requirements capture, market survey, specification of the tender/request for proposal documents, engaging with suppliers, evaluating proposals and managing the installation.
Duration: Half-day
Level: Managers/faculty
Sincerely,
Bilel Hadri, HPCSAUDI17 tutorial chair