The OneAPI Center will be directed by the University of Utah Extreme Data Management Analysis and Visualization Center (CEDMAV) Collaboration will include Lawrence Livermore National Laboratory Center for Applied Scientific Computing (CASC). will accelerate ZFP compression software Use oneAPIOpen standards-based programming on multiple architectures to advance Excel computing.
Participants said the center’s efforts expand the long-term collaboration of organizations dedicated to developing advanced data formats and layouts for efficient storage and providing access to large-scale scientific data for high-performance computing (HPC) architectures.
“The University of Utah CEDMAV, in collaboration with LLNL’s CASC, has pioneered research into managing extreme data applications that include scientific simulation and experimental facilities,” said Manish Prashar, director of the University of Utah’s Institute of Scientific Computing and Imaging. This collaboration has a proven track record in developing and publishing open source scientific software that finds wide adoption in the communities of interest. This API’s Center of Excellence will further this collaboration and help this academic research find practical adoption on multiple architectural systems. “
Developed by LLNL, ZFP is a state-of-the-art floating point data loss controlled compression program that has become a de-facto standard in the HPC community, with many scientific, engineering, and user applications. ZFP(de) compression is particularly amenable to parallel execution of data by decomposing it into small, independent data blocks, and parallel backends have been developed for the OpenMP, CUDA, and HIP programming paradigms, according to LLNL computer scientist Peter Lindstrom.
“As a pioneer in ZFP development, I am excited about this opportunity with our longstanding partners at the University of Utah to extend the capabilities of our ZFP compressor to work efficiently on next-generation supercomputers, including Argonne National Laboratory’s Aurora system, one of the first Exascale systems in the world,” Lindstrom said. world.” “The resulting compression program will allow for large-scale scientific computing applications, among other things, to effectively enhance memory capacity and bandwidth while significantly reducing connectivity, I/O time, and offline storage.”
With the ZFP development team at LLNL, the API Center of Excellence will develop a portable, scalable, performance-based SYCL-based ZFP interface that runs on accelerator architectures across different vendors, including Intel Corporation GPUs in the data center. As one of the software technologies selected by the Department of Energy (DOE) Exascale Computing Project (ECP), ZFP has been certified by large-scale parallel simulations and technologies running on some of the world’s largest supercomputers, which will benefit many high-definition scientific applications. Furthermore, ZFP’s widespread adoption in industry and academia will aid the development of many large-scale data management technologies, including HDF5, ADIOS, OpenZGY, OpenVisus, and Zarr.
The development of ZFP’s high-performance SYCL port on multi-vendor acceleration architectures will benefit many high-definition supercomputing applications and better showcase the power of an open, standards-based ecosystem.
“The work of the University of Utah and Lawrence Livermore National Laboratory in developing the SYCL-based high-performance ZFP library helps make large-scale scientific data available for high-performance computing architectures, enabling exascale applications to target multiple accelerator architectures,” said Scott Abeland, senior director. From Intel Developer Ecosystem Programs. “This latest Center of Excellence will showcase how open, standards-based oneAPI development benefits the developer community.”
CEDMAV’s research approach stems from a systematic assessment of the needs of HPC applications and how they lead to new and innovative investigation, followed by practical validation and dissemination to broader communities. CEDMAV’s previous collaborations with LLNL include joint research projects, dual-appointed staff, and student and postdoctoral interns.
“It is a great honor for CEDMAV to establish this API Center of Excellence in collaboration with LLNL. This will give a great opportunity to consolidate and expand our collaboration with the support and collaboration of Intel engineers,” said Valerio Pascucci, Founding Director of CEDMAV and former CASC Data Analysis Group Leader at LLNL Valerio Pascucci. “It is exciting to see the emergence of the oneAPI programming model that we plan to fully embrace in this project. In particular, SYCL’s cross-platform abstraction will greatly increase the productivity of our teams creating performance codes that will run efficiently on modern heterogeneous architectures. Diverse hardware and software architectures are becoming ubiquitous in high-performance systems, and oneAPI technology will greatly increase the impact of ZFP in a wide range of applications.”
CEDMAV at the University of Utah is internationally recognized for its activities that include theoretical and algorithmic research, systems development, and the deployment of tools for dealing with extreme data. This research lies at the intersection of scientific visualization, big data management, HPC, and data analytics.
The Center for Applied Scientific Computing serves as a window to the LLNL School more broadly in the computer science, computational physics, applied mathematics, and data science research communities. With academic, industrial, and other government laboratories partners, it conducts world-class scientific research and development on issues critical to national security.
oneAPI is an open, unified, multi-architecture programming model for CPUs and acceleration architectures (GPUs, FPGAs, etc.). Based on the standards, the programming paradigm simplifies software development and delivers unparalleled performance for accelerated computing without proprietary locking, while enabling existing code integration. With oneAPI, developers can choose the best architecture for the specific problem they are trying to solve without having to rewrite the program for the next architecture and platform.