TRAINING
Simulating complex physical systems and solving large problems requires computing power far beyond what can be achieved with a simple desktop computer. Moreover, with Big Data, the computational performance requirement is becoming more important day by day.
HPC (or High-Performance Computing) is thus becoming an essential tool for industry and research today and tomorrow. A large part of the training will be devoted to an implementation work on a virtual distributed infrastructure.
COURSE OBJECTIVES
• Presenting the fundamental principles and best practices of HPC computing on distributed architectures.
• Understanding the new challenges of HPC in the cloud.
• Putting knowledge into practice on well-known examples from the scientific and industrial world.
TARGET AUDIENCE
This training is intended for engineers and researchers likely to use HPC and wishing to acquire a first experience or cluster users looking to update their knowledge on the latest technologies.
INSTRUCTORS
The speakers are Artelys optimization engineers experienced in the use of HPC.
PROGRAM
Presentation of High-Performance Computing
• Main issues and need of parallelization.
• Application examples.
• Hardware and software components of High-Performance Computing (processors, memory, applications for the implementation of parallelization on the hardware).
• Performance measurement of calculations and improvement techniques.
• Introduction to parallelism performance indicators.
Architecture of HPC systems
• Shared memory architecture.
• Distributed memory architecture.
• Hybrid architectures.
High-performance architecture management systems: Scheduling and Load-balancing with a high-performance infrastructure management system.
• Performance metrics and monitoring.
• Overview of the main systems: SGE, SLURM, OpenPBS, Torque and Celery.
• Scalables cloud technologies: Container as a Service, K8S and Knative, Azure et AWS Batch, Parallel Cluster – GCP Cloud HPC Toolkit – Azure CycleCloud.
Setting up an HPC cluster:
• Setting up a local cluster with Docker.
• Launching first jobs.
• Installing apptainer.
• Building apptainer images.
• Launching containerized jobs.
PRACTICAL INFORMATION
Next session
06/12/2024 – 06/13/2024
Training Duration
2 days
Training cost
Inter-company: €1,800 excluding taxes per person
In-house: €9,000 excluding taxes per session
This includes the materials, meals, coffee, and handouts. Computers are provided.
Artelys is a training institution, registered in France under registration number 11754066975.
Intéressé par cette formation ? Inscrivez-vous !
