I GPU Computing Workshop

Data: 
30/10/2018 - 09:00 - 01/11/2018 - 17:30
Local: 
Auditório do IEE/USP (Av. Prof. Luciano Gualberto, 1289, Cidade Universitária)


Graphical processing units (GPUs) provide massive parallelism enabling one to perform computations on a basic computer that until very recently required a supercomputer. GPUs not only allow new kinds of problems to be tackled but are also the core technology used in many of the world’s fastest and most energy-efficient supercomputers. For instance, the top ten supercomputers in the world changed from being 100% based on CPUs to GPU systems. In this workshop, we will have a series of talks exploring the game-changing technology of GPU-based parallel computing.

Invited Talks
 
• Seismic Data Structure Detection With Deep Learning. Edson Borin (UNICAMP)
 
• A GPU Approach for the Phase Problem. Eduardo Miqueles (LNLS)
 
• Portability and Efficiency of Production Codes at Petrobras. Jairo Panetta (ITA)
 
• Powering the World's Fastest Supercomputers with GPUs. João Paulo Navarro (NVIDIA)
 
• Applied Models on GPU Supercomputers. Murilo Boratto (CIMATEC)
 
• HPC and Atmospheric Sciences. Pedro Leite da Silva Dias (IAG USP)
 
• Machine Learning and Deep Learning for Oil & Gas. Pedro Mário Cruz e Silva (NVIDIA)
 
• GPUs in Earth Sciences: There is NO Free Lunch! Pedro Paes Lopes (Exaflop)
 
• Improving Scientific Computing in Space Physics with GPU Solutions. Reinaldo Rosa (INPE)
 
• Accelerating Astrophysical Simulations with GPUs. Rodrigo Nemmen (IAG USP)