Projects
ECP ADIOS2
Department of Energy (DoE) project, a widely used software among the world's fastest supercomputers. ADIOS2 is an adaptable input/output system designed to support exascale computing applications.
Role: Tech lead, Release engineer, Core developer
Responsibilities: In charge of deployments in current fastest supercomputers (Frontier/Summit/Crusher).
Viskores
Visualization plugin for streaming and working with large simulations data. Designed to work with ParaView, this project enhances the ability to handle and visualize massive datasets generated by scientific simulations.
Role: Project Lead and Core Developer
Technologies: C++, ParaView, ADIOS2, HPC
ParaView
World leading OSS computer visualization suite. ParaView is an open-source, multi-platform data analysis and visualization application designed to handle large datasets.
Role: Core Developer, Release Engineer
VTKm
HPC port of the world leading OSS visualization toolkit (VTK). VTK-m is a toolkit of scientific visualization algorithms for emerging processor architectures, designed for exascale computing environments.
Role: Core developer, Release Engineer
Responsibilities: In charge of deployments in current world fastest supercomputers.
BeeGFS
World leading performance oriented Distributed File System. BeeGFS is a parallel file system designed for high-performance computing environments.
Role: Core Developer
Contributions: Implemented features in both user-space and kernel-space. Ported its in-house RDMA network library to Omni-Path.
EclipseMR
Distributed and Parallel Task Processing with Consistent Hashing. A novel MapReduce framework prototype that efficiently utilizes a large distributed memory in cluster environments.
Role: Main author and developer
Publication: IEEE Cluster 2017
Velox Big Data Framework
Big data processing and storage framework based on EclipseMR, which is design on the foundations of a Distributed File system that can intelligently adapt to changing workload in distributed systems, efficiently utilizing all the resources of the system.
Role: Creator and primary maintainer
FactoryStreaming
Big data solution to analyze, store and visualize real-time data from Smart factories. It uses technologies such as stream processing, data pipelining, time series databases and industrial protocols.
Role: Architect