PDC Summer School 2024

The school focuses on the skills needed to utilise high-performance computing (HPC) resources for research. It consists of lectures and guided hands-on lab exercises using the Dardel  HPC system at PDC. The school includes an introduction to parallel algorithms, parallel programming, modern HPC architectures, performance analysis and engineering, and software engineering. Case studies in various scientific disciplines will be used to help illustrate practical research applications of the topics.

Dates

• 16 August 2024: a preliminary online introduction to Dardel from 09.00-11.00
• 19-24 August 2024: in person at KTH in Stockholm

Location

KTH Royal Institute of Technology main campus, Stockholm, Sweden

The lectures and hands-on exercises will be held in auditorium E2, Osquars backe 2 , Stockholm.

Number of places available

The number of places available in the summer school is limited to 50.

Who can attend

Ph.D. students and researchers in scientific and/or high-performance computing (with academic or industrial/commercial backgrounds) from anywhere in the world are welcome to apply to attend the school.

What you will learn

This school provides the knowledge and skills needed to utilize high-performance computing (HPC) resources for research purposes. Information about the topics that are covered in the school can be found in these detailed descriptions . The school is especially suited to computational and data scientists, and a variety of relevant HPC use cases will be studied to help illustrate the topics. The school consists of both lectures and guided hands-on lab sessions.

By the end of this school, participants will have done and/or will be able to do the following.

• Describe the architecture of modern supercomputers and the different computing units, including multi-core processors, NVIDIA graphics processing units (GPUs), and AMD GPUs
• List the sustainability aspects of HPC and its impact on the environment and society
• Use programming models for shared-memory (OpenMP) and distributed-memory (MPI) programming to develop scientific applications for supercomputers
• List the fundamental principles of programming NVIDIA and AMD GPUs and develop efficient GPU-accelerated applications
• Apply software engineering principles to develop scientific applications for supercomputers
• Use performance monitoring tools to identify performance bottlenecks and optimize application performance on supercomputers
• Apply high-performance data analysis and visualization techniques to scientific applications
• Practice programming on the PDC HPC supercomputers and apply the concepts that have been learned to real-world problems

Information for PhD students

A KTH doctoral course FDD3260: “High-Performance Computing for Computational Scientists” (see www.kth.se/student/kurser/kurs/FDD3260?l=en  for details) is associated with the PDC Summer School. If you are a PhD student and you attend the summer school, you are eligible to also enroll in the doctoral course FDD3260. To register for the doctoral course, choose that option in the form when you apply to attend the school. NOTE that master's students are not eligible to enroll in the doctoral course.  

For information concerning the doctoral course, including the contents, method of examination and more, refer to the doctoral course page ( www.kth.se/student/kurser/kurs/FDD3260?l=en ) and/or contact the course coordinator (whose details are available on the course page).

Requirements

Applicants must be able to communicate in English, and have previous programming experience in C/C++ or Fortran. Applicants need to be familiar enough with at least one of these languages to be able to write and execute programs. Basic knowledge of mathematical analysis and linear algebra, and also UNIX/Linux is required.

Applicants who need to refresh their UNIX/Linux skills are recommended to review this introduction to the UNIX shell , which provide an overview of what participants are expected to know before attending the summer school, or this introduction to basic to Linux for new HPC users .

Participants will need to bring a laptop for participating in the lab exercises.

How to apply

To apply to attend the school, fill in the application form .

Deadline for applications: 10 May 2024

Note that SeSE doctoral students are Ph.D. students attending one of the SeRC or eSSENCE partner universities in Sweden (namely LU, LiU, KTH, KI, SU, UU and UmU).

Notifications of acceptance: PDC plans to send notifications of acceptance by the 15^th of May 2024.

Deadline for payment of fees: 30 May 2024

Fees

SeSE Ph.D. student: 1,000 SEK
SeSE doctoral students are Ph.D. students attending one of the SeRC or eSSENCE partner universities in Sweden (namely LU, LiU, KTH, KI, SU, UU and UmU).

Ph.D. students and researchers from other public research institutions (in Sweden or outside Sweden): 5,000 SEK

Industry/business researcher: 15,000 SEK

Contact

If you need further information or help in relation to the PDC summer school, please contact hpc-summerschool@pdc.kth.se .

Schedule

In general, there will be lectures alternating with hands-on lab sessions from 8:30-12:00 and 13:15-17:15 most days, followed by guided tours of Dardel for small groups. A detailed timetable will be published soon.

Speakers

The following are some of the speakers who will give lectures during the summer school.

• Ana Lucia Varbanescu (Associate Professor, University of Twente and University of Amsterdam, Netherlands)
• Erwin Laure (Director of Max Planck Computing and Data Facility, Munich, Germany)
• Niclas Jansson (Researcher, PDC, KTH Royal Institute of Technology, Sweden)
• Radovan Bast (Senior Engineer, The Arctic University of Norway)
• Stefano Markidis (Professor of Computer Science, KTH Royal Institute of Technology, Sweden)
• Szilárd Páll (Research Engineer, PDC, KTH Royal Institute of Technology, Sweden)
• Xin Li (Researcher, PDC, KTH Royal Institute of Technology, Sweden)

Organisers