Paulo Souza
About Paulo Souza
Paulo Souza is a Principal Software Engineer at Core Scientific, where he has worked since 2020. He has extensive experience in high-performance computing (HPC) solutions, particularly in the oil and gas industry, and has held various consulting roles in Brazil over the past two decades.
Work at Core Scientific
Paulo Souza has been serving as a Principal Software Engineer at Core Scientific since 2020. In this role, he focuses on developing advanced software solutions that leverage high-performance computing (HPC) technologies. His work contributes to the company's initiatives in the blockchain and AI sectors, utilizing his extensive background in HPC to enhance system performance and reliability.
Education and Expertise
Paulo Souza studied Computer Engineering at Pontifícia Universidade Católica do Paraná, where he earned his Bachelor's degree from 1995 to 1999. He furthered his education by obtaining a Master's degree in Computer Science from Pontifícia Universidade Católica do Rio de Janeiro in 2015. His academic background supports his expertise in high-performance computing, particularly in the oil and gas industry, where he specializes in hybrid HPC and HPC containers.
Background
Paulo Souza has a diverse professional background in high-performance computing. He began his career as a Trainee at Pontifícia Universidade Católica do Paraná from 1995 to 1998. He has held various consulting positions, including roles at Petrobras, where he worked for 13 years as an HPC Consultant and Software Engineer. His experience also includes positions at Ontropos, Inc., RStor, and Atrio Inc., where he focused on HPC solutions and optimization.
Achievements
Throughout his career, Paulo Souza has developed portable fault-tolerant MPI applications capable of operating over thousands of cores. He has extensive experience in porting seismic imaging applications to CUDA, achieving significant performance improvements. His expertise also includes optimizing x86 code using gcc inline SIMD assembly programming and vectorization, as well as programming for the IBM Cell SPE with a PS3 cluster.