PAX-HPC General Training on GPU programming with Directives and Tailored Session
The training event will start with a tailored session for PAX-HPC attendees only on Wednesday 12 March when training/consultancy will be provided on CUDA Fortran and GPU aware MPI as people requested previously. We will then have a 2-day PAX-HPC and ARCHER2 joint training course on the GPU programming with directives using OpenMP and OpenACC (covering C/C++/Fortran) on Thursday 13 March and Friday 14 March.
Please note the locations are different for the first afternoon (12 March PM) and the following 2-day training course (13-14 March).
We would recommend all PAX-HPC attendees to complete your registration by Monday 10 February if possible. The registration will then be opened to the wider ARCHER2 users.
12-14 March
– 12 March (PM): Room 2.55, Bayes Centre, The University of Edinburgh, 47 Potterrow, Edinburgh, EH8 9BT
– 13 – 14 March: Room 4.20, Charteris Land, The University of Edinburgh, Holyrood Road, Edinburgh, EH8 8AQ
Registration link: https://forms.office.com/e/8u4ge7ndsN?origin=lprLink
Registration deadline, recommended for PAX-HPC attendees: 10 February 2025
PAX-HPC Scientific Committee Meeting
14:00 on 13th January, finish 15:00 on 14th January 2025
- Building: Nancy Rothwell Building, https://www.manchester.ac.uk/about/maps/interactive-map/?id=540 (large black building, #24 on the university maps). This is a 10-to-15-minute walk from the main train stations of Manchester Piccadilly or Oxford Road.
- Room: 3A.012 (Core 1, Floor 3)
| Time | Activity | Presenter |
| 14:00 – 14:15 | Welcome to Manchester | Benedict D. Rogers |
| 14:15 – 15:15 | Achievements of PAX-HPC 1: (1) Accelerating SWIFT’s Task-based parallelism with GPUs: How close are we to exascale? (2) SWIFT 2: Towards Exascale SPH with Task Parallelism and the Peano4 Framework Engine (3) Complex multiscale workflows and parallel I/O |
Abouzied Nasar
Mladen Ivkovic Rajany Kv |
| 15:15 –15:35 | Tea & Coffee break | |
| 15:35 – 16:35 | Achievements of PAX-HPC 2: (1) A new hybrid engine for electron repulsion integrals in CP2K (2) Unbiased Monte Carlo Simulations of Li Battery Cathode Materials (3) Quantifying Uncertainty in DFT-Driven Structure Prediction |
Sergey Chulkov
Scott Woodley Alexey Sokol |
| 16:35 –16:45 | Comfort break | |
| 16:45 – 17:45 | Guest Speakers: A look forward in HPC with AMD Planning and co-design process for the next generation of DiRAC service |
Maggie Anderson/Rowan Pavlovic
Mark I. Wilkinson |
| 18:30 – 21:00 | Evening Meal at Christie’s |
DAY 2: PAX-HPC Status Quo & Future Challenges
| Time | Activity | Presenter |
| 9:00 – 9:30 | Invited Keynote Lecture: Breaking the Million-Electron and 1 EFLOP/s Barriers Winner of the 2024 Gordon Bell Prize |
Prof. Giuseppe Barca (University of Melbourne) |
| 9:30 – 11:00 | Achievements of PAX-HPC 3 (cont): (1) Acceleration of Electronic Structure Codes on Heterogeneous Hardware (2) CASINO and ONETEP GPU port updates. Fun and games with FORTRAN and OpenACC (3) Accelerating CASTEP and GPUs |
Marcello Puligheddu
Ben Thorpe Matt Smith |
| 11:00 – 11:30 | Coffee & Tea break | |
| 11:30 – 12:30 | Guest Speaker: Energy Efficient Computing Through Mixed-Precision Algorithms and Floating-Point Emulation Training: “What’s next for our exascale developments?” (5 mins summary & 10 minutes on future needs and involvement) Knowledge Transfer: Where next? |
Filippo Spiga (NVIDIA)
Xu Guo Phil Hasnip |
| 12:30 – 13:30 | Lunch | |
| 13:30 – 15:00 | PAX-HPC Discussion: Outcomes, Lessons Learnt and Future Needs | Chair: Benedict Rogers |
| 15:00 | Finish |
PAX-HPC General Training on Advanced Parallel I/O and Single Node Optimisation
This event will include a 1-day training on Advanced Parallel I/O, and a 2-day training of Single Node Optimisation (including memory optimisation) as selected by PAX-HPC members, followed by the final day of hackathon when attendees will be able to work together on their codes using the optimisation techniques introduced during the training. EPCC staff will be available to give individual hands-on advice.
- Monday 11 November: 09:30 – 17:00 Advanced Parallel I/O
- Tuesday 12 November: 09:30 – 17:00 Single Node Optimisation Day 1 (joint training day with ARCHER2)
- Wednesday 13 November: 09:30 – 17:00 Single Node Optimisation Day 2 (joint training day with ARCHER2)
- Thursday 14 November: 09:30 – 16:00 Hackathon & wrap-up discussion
Please note the Single Node Optimisation will be a joint training with the ARCHER2 training. The Advanced Parallel I/O training day and the hackathon day will be exclusive to the PAX-HPC attendees only.
Room G.03, Bayes Centre, The University of Edinburgh, 47 Potterrow, Edinburgh, EH8 9BT
11-14 November 2024
PAX-HPC General Training on Performance Analysis and Debugging
This event will include training on performance analysis and debugging tools on ARCHER2/Cirrus followed by a hackathon when attendees will be able to work together on their codes using the tools. The main aim is to help attendees to better understand the behaviour and performance of their own codes
Room G.03, Bayes Centre, The University of Edinburgh, 47 Potterrow, Edinburgh, EH8 9BT
10-13 June 2024
Registration deadline: Friday 10 June 2024
PAX-HPC Scientific Committee Meeting
This meeting is by invitation only for the members of the Excalibur PAX HPC project
22nd – 23rd April 2024
Private Dining Room – Lancaster University
Hosted by Neil Drummond – Lancaster University
Programme
| Monday 22nd April 2024 | |
|---|---|
| 11.00 am | Arrival / tea / coffee |
| 11.45 am | Welcome (Neil Drummond) |
| 11.50 am | Talk: Acceleration of Electronic Structure Codes on Heterogeneous Hardware (Marcello Puligheddu) |
| 12.30 pm | Talk: Application of a Master–Worker Parallelism on Materials Modelling (Woongkyu Jee) |
| 1.00 pm | Lunch |
| 2.00 pm | Talk: SWIFT2 (Mladen Ivkovic) |
| 2.40 pm | Talk: Accelerating CASINO using GPUs (Ben Thorpe and Neil Drummond) |
| 3.20 pm | Tea / coffee |
| 3.40 pm | Talk: Attempts at Particles Without MPI and the Hardware Landscape (Alastair Basden) |
| 4.20 pm | Talk: CASTEP for Exascale (Matt Smith and Phil Hasnip) |
| 5.00 pm | Talk: Yet Another HPC Benchmarking Project (Rob Welch) |
| 5.40 pm | Discussion on training requirements (led by Xu Guo) |
| 7.00 pm | Dinner |
| Tuesday 23rd April 2024 | |
|---|---|
| 9.30 am | Annual State of the Project address (Scott Woodley) |
| 10.00 am | RSEs: Future funding, etc. (led by Alexey Sokol) |
| 11.00 am | Tea / coffee |
| 11.30 am | Plans for PAX-HPC review paper (led by Scott Woodley) |
| 1.00 pm | Lunch |
| 2.00 pm | Knowledge exchange (led by Phil Hasnip) |
| 3.30 pm | Tea / coffee |
| 4.00 pm | Finish / (Neil Drummond) |
Training: PAX-HPC General Training on Parallel I/O and GPU Programming Models
Dates: 17th-19th January 2024
Location: Room G.03, Bayes Centre, The University of Edinburgh, 47 Potterrow, Edinburgh, EH8 9BT
Registration: https://bit.ly/3SMZ18e
PAX Project Meeting
This meeting is by invitation only for the members of the ExCALIBUR PAX-HPC project.
Tuesday 16 May – Wednesday 17 May 2023
Scott Logic lecture theatre (Computer Science building), University of Durham
Hosted by Tobias Weinzierl
| Tue, 13:30-14:00 | Welcome | Consortium members arrive, coffee
Welcome note |
Tobias Weinzierl |
| Tue, 14:00-15:30 | Status update | – GPU Acceleration for SPH With Inter-dependent Fine-grained Tasking – CASTEP for Exascale HPC – Tasking and asynchronous MPI in SWIFT – Uncertainty quantification on materials science |
– Abouzied (Manchester)
– Matt (York) – Peter and Alastair (Durham, ICC) |
| Tue, 15:30-16:00 | Coffee break | ||
| Tue, 16:00-17:30 | Status update | – Swift 2: keeping the good, discussing the bad, removing the ugly – A C++ language extension to write memory-efficient Lagrangian codes – Recent activities in QEVEC |
– Mladen
– Pawel (Durham, CS) – Viv (Strathclyde) |
| Tue, 17:30-18:30 | Discussion | Research data management | Ben Rogers |
| Wed, 9:00-10:30 | Breakout sessions | Coffee
– Room 1: How do we use non-blocking MPI and MPI’s new features such as RMA? |
Participants join group of interest and discuss their algorithmic approach. Goal is to identify best practices and common pain. |
| Wed, 10:30-11:00 | Coffee break | ||
| Wed, 11:00-12:00 | Update | Reps of breakout sessions meet and introduce outcomes to whole consortium | Whole consortium |
| 12:00-13:00 | Light lunch | ||
| Wed, 13:00-14:30 | Discussion | – Upcoming events and conference attendance – Presence at CIUK – Joint publications |
Whole consortium |
| Wed, 14:30-15:30 | Wrap-up | Summary, next steps, next meeting | Scott Woodley, Tobias Weinzierl |
| Wed, 17:30-19:00 (est) | Project management | Project management meets | (panel members only) |
| Wed, 17:30-19:00 (est) | Homepage hackathon | Hackathon to create joint project webpage and dissemination material such as posters (to be sent to EPSRC afterwards) | All RSEs and PDRAs invited |
Cross-Cutting Update July 2021 and Future call
12pm Midday (BST) Tuesday 6 July 2021
4 Cross-cutting PIs (Matthew Watkins, Viv Kendon, Peter Coveney, Ilian Todorov) will give 5-minute interview intros, likewise Rob Bowler will talk on experiences regarding “Practicalities of task based parallelisation – experiences so far?”.
MMM DDWG Scientific Case Studies discussion
Monday 24th May 2021
Intel: OneAPI: A Unified, Standards-Based Programming Model
Tuesday 10th November 2020
ExCALIBUR – An algorithmic approach to Exascale design
Richard Gunn and Jeremy Yates
Wednesday 10th June, 11:00 BST