The PRACE Autumn School 2021 will bring together undergraduate and graduate students and postdoctoral researchers in different disciplines of scientific computing.
The course consists of lectures and hands-on training on modern, GPU-accelerated high-performance computing: GPU programming and GPU code optimization at scale, as well as understanding and applying machine learning methods. The course includes also scientific case studies about using GPUs in various disciplines. The tutors and lecturers are experts in these fields providing various interesting aspects in the course topics.
The five days of school (11-15 October 2021), organised by CSC - IT Center for Science, are filled with hacking, leisure activities and having good time together in Vuokatti, Finland. Also, the Autumn School includes a visit to the CSC’s Data Center in Kajaani, the home of one of the most powerful supercomputers in the world: LUMI. It is a unique opportunity for the participants to familiarize themselves with the cutting-edge data infrastructure in the field of HPC.
The applicants are expected to select according to their interest one of the three parallel hands-on tracks they wish to attend :
- Introduction to deep learning
- Introduction to GPU programming with OpenMP
- Hackathon: Optimizing GPU applications on LUMI supercomputer.
Maximum of 40 participants will be selected from registered applicants to attend the event. In order to become selected, the applicant should meet the required background skills for the Hands-on Track they apply for. The decisions of acceptance or rejection will be done in the order of arriving applications. Background and experience requirements for each of the Hands-on Tracks can be found on the section 2 on this page.
Selected applicants are expected to confirm his/ her attendance as soon as possible after been selected, but latest by 31. August 2021, 12:00 CEST. More detailed instructions of confirming the attendance will be provided to the selected applicants by email after the decision on selection.
1. Learning Outcome
After the course the participants should be familiar with modern GPU-accelerated supercomputers and the possibilities they offer. Depending on the chosen hands-on track the participants will have
Skills and knowledge needed to begin applying deep learning for different tasks and utilizing the GPU resources for training and deploying their own neural networks.
Basic skills needed in utilizing OpenMP offloading with new or existing programs
- Basic skills to optimize existing GPU-applications at scale
2. Background skills
Prerequisities depend on the chosen hands-on track.
2.1 Introduction to deep learning
The participants are assumed to have a working knowledge of the Python programming language. Previous experience in machine learning is not required, but knowledge of the fundamentals can be beneficial. Basic knowledge of a Linux/Unix environment will be assumed.
2.2 Introduction to GPU programming with OpenMP
The participants are assumed to have a working knowledge of Fortran and/or C programming languages. In addition, fluent operation in a Linux/Unix environment will be assumed.
2.3 Hackathon: Optimizing GPU applications on LUMI supercomputer
The participants need to have their own application code to work with during the hackathon, and have a working knowledge of developing the application. In addition, familiarity with some performance analysis tools is expected.
3. Selection of attendees
The selected applicants are expected to confirm their attendance as soon as possible after been selected but at latest by 31 August 2021, 12:00 CEST. The confirmation from participants will be binding.
The selection bases on the background knowledge and experience meeting the course prerequisities. Also geographical distribution is considered.
The decisions for acceptance will be made in the order of arriving applications. Notification about acceptance or rejection or inclusion in the waiting list will be sent to the applicants within five business days after submission of the applications.
Please note that applicants will receive an automatic confirmation of registering the application but the applicant will still receive a separate email confirming whether he/ she has been selected to attend the PRACE Autumn School in Finland.
More detailed instructions on confirming the attendance will be provided to the selected applicants by email after the decision on selection.
4. Costs and other practicalities
The course will be primarily held onsite. In case of covid-19 restrictions, lectures can be held as a hybrid event. Hands-on tracks will not be held online. In case of cancellation of the onsite event due to covid-19 situation, the full event will most likely be cancelled.
What the PRACE Autumn School covers:
- lessons and hands-on training
- catering (lunches and dinners)
- onsite visit to Kajaani Data Center
- bus transportation option from Kajaani to Vuokatti and from Vuokatti to Kajaani (the exact timing of the bus departure will be communicated to the selected attendees).
- Please note that if you travel by another bus than by the bus arranged by the event organisers, the travel is by participant's own expense.
Accommodation, travels and recreation activities are on participants’ own expense.
We kindly recommend the attendees to include cancellation policy on their travel reservations. The participants are also expected to cancel the reservations for room booking and travels themselves in case of covid-19 situation causing the cancellation of the event.
One night at the Vuokatti Sport Resort costs 75 euros including breakfast (all 4 nights during the event cost 300e). Instructions for reserving the accommodation onsite will be provided to the selected participants. The participants are expected to make the room reservations by themselves.
Concerning further practicalities and recreation activities, the selected applicants will be contacted by the Autumn School organizers later on by email.
For more information click "LINK TO ORIGINAL" below.