 |
||
|
||
Home > TERATEC FORUM >Workshops
Workshop 05 - 09:00 am to 12:30 am
Is the Cloud the (inevitable) future of HPC?
Urban legends and realities
Chaired by Gilles Tourpe, Business Development Manager HPC, AWS and Xavier Vigouroux, Head of business development, CGG
Running Scientific Applications on a Cloud-based Environment: Challenges and Opportunities
By Benjamin Pajot, Field Support Manager for the AMD HPC Centre of Excellence and José Noudohouenou, Sr. Staff Software Engineer, AMD
Demands in Cloud-based HPC and AI environments continue to grow and nowadays, various offers exist including the ones from general cloud service providers as well as offers from other industries like semiconductors. Even though many challenges in working in a cloud environment still exist, scientists have today the choice between on-prem and on cloud computing.
This talk focuses on key factors to consider when making the decision on moving from on-prem to cloud and how to seamlessly make the transition. Also, an example of cloud-based computing services provided by a semiconductor is exposed.
 |
Biography: After studying Modelling and Scientific Computing, I dedicated my PhD to develop and implement innovative numerical methods applied to the Atmospheric Chemistry in CERFACS. I kept up my journey in the HPC world working 2 years on Seismic Imaging for the Seiscope Consortium before joining BULL/Atos. This adventure will last 7 years during which I will cover many aspects of the life of a cluster, from its design to its delivery, production phase and support. I am now working for the AMD Center of Excellence, guiding our Partners and Customers through our HPC offer and supporting our Engineer and Sales teams with a deep level of HPC application and system level expertise. |
 |
Biography: Jose Noudohouenou held a Ph.D. from University of Versailles(France) in 2013 and is specialized in application characterization, code optimization and performance prediction. Then, he worked at Exascale Computing Research Lab (France) as a postdoctoral researcher focusing on HW/SW codesign issues and the problem of matching codelet equivalence. While continue tuning scientific applications, he has accepted a position at Intel Corporation(USA), and later at Intel Federal where he was involved in various Federal projects including porting and optimizing HPC applications onto Intel GPUs. Jose joined AMD in 2021 and continues optimizing both HPC and AI applications onto AMD GPU while he is the AMD Center of Excellence (CoE) lead for Adastra machine at the CINES HPC center in Montpellier, France. |
Register now and get your badge here >>>