6th KAUST-NVIDIA Workshop on "Accelerating Scientific Applications using GPUs"

Summary of agenda




Saber Feki, Workshop Chair            Bilel Hadri and Nagarajan Kathiresan, Workshop Co-Chairs

Accelerating Scientific Applications Using GPUs

6th KAUST-NVIDIA Workshop

The KAUST Supercomputing Core Laboratory (KSL) in collaboration with NVIDIA are pleased to invite you to the 6th annual workshop on "Accelerating Scientific Applications using GPUs" that provide an opportunity to learn about the latest advances by NVIDIA for Scientific Computing and Artificial Intelligence as well as showcasing the related KAUST success stories using GPUs.  


     March 12, 2019

     Time: 9:00AM to 4:00PM

     Lecture Hall 2 room 2325 , bldg.9


08:30 9:00

Breakfast and Registration

09:00 09:10


Dr. Jysoo Lee, Facilities Director, Research Computing Core Labs &

Mr.Piero Altoe, Business Development Manager, NVIDIA

09:10 10:10

Keynote Talk #1: An Accelerated Computing System is Different

Dr.Timothy Lanfear, Director, Solution Architecture and Engineering EMEA, NVIDIA

10:10 10:30

GPU Resources at KAUST: the present and the future

Dr.Saber Feki, Computational Scientist Lead, Supercomputing Core Laboratory, KAUST

10:30 – 11:00

Coffee Break and Group Photo

11:00 –11:30

SwitchML: Scaling Distributed Machine Learning with In-Network Aggregation

Dr. Marco Canini, Assistant Professor, Extreme Computing Research Center, KAUST

11:30 – 12:00

A Five-Year Status of GPU-accelerated Applications at ECRC

Dr.Hatem Ltaief, Senior Research Scientist, Extreme Computing Research Center, KAUST

12:00 – 13:30


13:30 – 14:15

Keynote Talk #2: OpenACC, GPU Programming and Parallelware Tools

Dr. Manuel Arenaz, CEO at Appentra Solutions and professor of computer science, University of a Coruña, Spain

14:15 – 14:45

High-performance Asynchronous Execution of the Reverse Time Migration for the Oil and Gas Industry

Ms.Amani Alonazi, PhD Student, Computer Science, KAUST

14:45 15:15

Matrix Algebra Framework for Portable, Scalable and Efficient Query Engines for RDF Graphs

Mr.Fuad T. Jamour, PhD Student and Prof. Panagiotis Kalnis, Computer Science, CEMSC, KAUST

15:15 15:30

Coffee Break

15:30 – 16:00

Leveraging ExaGeoStat with Mixed Precision Algorithms for Modeling and Predictions in Geo-Statistics Applications

Dr.Sameh M. Abdulah, Dr.Hatem Ltaief, Dr.Ying Sun, Dr.Marc Genton, and Dr.David Keyes, Extreme Computing Research Center, KAUST

16:00 – 16:15

GPU Hackathon

Dr.Bilel Hadri, Hackathon Chair and Computational Scientist, Supercomputing Core Laboratory, KAUST

16:15 – 16:30

Closing Remarks 

Dr. Saber Feki, Workshop Chair and Computational Scientist Lead, Supercomputing Core Laboratory, KAUST


     Registration is free but seats are limited, therefore please register here.

Keynote speaker


Timothy Lanfear manages the European solution architecture and engineering team in NVIDIA’s Enterprise Solutions Group. He has twenty-five years’ experience in HPC, starting as a computational scientist in British Aerospace’s corporate research centre, and then moving to technical pre-sales roles with Hitachi, ClearSpeed, and most recently NVIDIA. He has a degree in Electrical Engineering and a PhD for research in the field of graph theory, both from Imperial College London. He is a Chartered Engineer and Member of the Institution of Engineering and Technology.


  An Accelerated Computing System is Different
The title of the presentation is a quotation from Steve Oberlin, the CTO of NVIDIA’s accelerated computing business unit. An accelerated computing system should not be viewed as another type of computer to target for porting legacy applications, but rather a new computational platform enabling new approaches to simulation and modeling. This change was proven by the 2018 Gordon Bell Prize awards in which new algorithms and coupling AI methods with traditional simulation have advanced computational science. NVIDIA is innovating across all layers of the solution stack: the hardware and processor architecture; the programming model, libraries, tools, compilers, and new approaches to simulation.

Training Sessions

Prior to the workshop and hack-a-thon, two training sessions are scheduled in March 10-11 on untitled "Deep Learning Demystified" and "OpenACC Programming" by Nvidia and Appentra respectively

Training #1: Deep Learning Demystified


        March 10, 2019
        Time: 9:30AM to 4:30PM
        Lecture Hall 2 room 2325 , bldg.9 

Trainer Bio-data:

Piero Altoè

Piero Altoè is responsible of the HPC segment in Middle East at NVIDIA and he brings more than 8 years of experience in computational chemistry. he is author of 30 papers published on international scientific journals. He joined the NVIDIA in 2016. Piero Altoè achieved his Ph.D. in 2007 at Università di Bologna focusing his attention on the development and applications of hybrid quantum and molecular mechanics methods. Since 2002 he is involved in high performance computing activities related to software development and system administration.


9:30 AM



9:30 - 12:00 PM

LECTURE:  Deep Learning Demystified


The last five years has seen an unprecedented progress in Artificial Intelligence research. Neural Networks, which were perceived as a failed experiment of Machine Learning research, not only are successfully solving a wide range of everyday problems but in many cases significantly exceed human performance. The goal of this talk is to help to understand which problems can be solved with Deep Neural Networks today.

12:00 - 1.30pm



1:30 - 5:00 PM

Hands-on Course: Fundamental of deep learning for computer vision (Limited Seating)


In this hands-on course, you will learn the basics of deep learning by training and deploying neural networks. You will:

  Implement common deep learning workflows such as Image Classification and Object Detection.

  Experiment with data, training parameters, network structure, and other strategies to increase performance and capability.

  Deploy your networks to start solving real-world problems.

On completion of this course, you will be able to start solving your own problems with deep learning. 



Hands-on Training #2: OpenACC Programming


        March 11, 2019
        Time: 9:00AM to 5:30PM
        Library, Computer Lab.

Trainer Bio-data:

Manuel Arenaz is CEO at Appentra Solutions and professor of computer science at the University of A Coruña (Spain). Holds a PhD on advanced compiler techniques for automatic parallelization of scientific codes. After 10+ years teaching parallel programming at undergraduate and PhD levels, he strongly believes that the next generation of STEM engineers needs to be educated in HPC technologies to address the digital revolution challenge. Recently, he co-founded Appentra Solutions to commercialize products and services that take advantage of Parallware, a new technology for semantic analysis of scientific HPC codes. Parallware Trainer by Appentra is a step forward to enable experiential self-learning of parallel programming, one of the biggest technical challenges today. For this purpose, he led the foundation of the startup, conducted the technology transfer agreement and attended training courses for entrepreneur scientists. Now launching Parallelware Trainer to the HPC market.


9:00 - 9:15am



9:15 - 9:45 am


       An introduction to OpenACC and GPUs

9.45 - 10:30 am


      Accelerating with OpenACC

10:30 - 11:00am



11:00 - 11:30 am


      Parallel Patterns

11:30 - 12:00 pm


      Parallelizing the calculation of PI

12:00 - 1.30pm



1:30 - 3:00 pm


     OpenACC directives for improved performance

3:00 -  3:30 pm


3:30 - 4:00 pm


     Effectively managing data transfers

4:00 - 5:30 pm


      Parallelizing the LULESH microkernel





The workshop will be followed by a two-day GPU hack-a-thon where 5-6 teams of developers will be selected to port and accelerate their domain science application to GPU accelerators. These teams will be guided by OpenACC and CUDA mentors from NVIDIA and KAUST.


        March 13-14, 2019
        Time: 9:00AM to 4:00PM
        Library, Computer Lab


        Hack-a-thon proposal is free, limited to 5-6 teams, therefore please submit your proposal here.


        Nvidia will give Two GTX1080ti for winners


OpenACC OpenACC web page
OpenACC reference guide
OpenACC programming guide
OpenACC getting started guide
CUDA Cuda with C/C++
Cuda with Fortran
GPU Libraries GPUs libraries
Other Matlab and GPU

Reference from last year: 

      GPU Hack-a-thon, 2018 materials: https://github.com/KAUST-KSL/GPUHACKATHON18

KAUST Map/Locations


and please contact us at training@hpc.kaust.edu.sa if you need further information.