The University of Utah has announced the creation of a new oneAPI Center of Excellence focused on developing portable, scalable, and high-performance data compression techniques.
The oneAPI center will be directed out of the University of Utah Center for Extreme Data Management Analysis and Visualization (CEDMAV) and will involve cooperation from the Center for Applied Scientific Computing (CASC) of Lawrence Livermore National Laboratory. It will accelerate ZFP compression software using oneAPI’s open, standards-based programming across multiple architectures to advance exascale processing.
Attendees said the center’s efforts extend long-standing collaborations 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 the LLNL CASC, has pioneered research in managing extreme data applications involving scientific simulations and experimental structures,” said Manish Prashar, Director of Scientific Computing and Imaging Institute of the University of Utah. “This collaboration has a long experience in the development and distribution of scientific open source software that finds wide adoption in the communities of interest. This oneAPI Center of Excellence will strengthen this collaboration and help this academic research find practical adoption on multi-architecture systems. “
Developed by LLNL, ZFP is state-of-the-art software for lossless and error-controlled floating point data compression that is becoming a de facto standard in the HPC community, with numerous users and scientific and engineering applications. ZFP (de) compression is particularly suited to parallel data execution by breaking it down into small, independent data blocks, and parallel backends for OpenMP, CUDA and HIP programming models have been developed, according to computer scientist Peter Lindstrom of LLNL .
“As ZFP Development Manager, I am excited about this opportunity with our longtime collaborators at the University of Utah to extend the capabilities of our ZFP compressor to run efficiently on next generation supercomputers, including the system. Aurora of Argonne National Laboratory, one of the world’s first exascale systems, “said Lindstrom.” The resulting compression software will enable large-scale scientific computing applications, among others, to effectively increase memory capacity and bandwidth, significantly reducing communication and I / O times and offline storage. “
With LLNL’s ZFP development team, the oneAPI center of excellence will develop a portable, scalable and high-performance SYCL-based ZFP backend that runs on acceleration architectures from multiple vendors, including Intel data center GPUs. As one of the software technologies selected by the Department of Energy’s (DOE) Exascale Computing Project (ECP), ZFP is being adopted by massively parallel simulations and technologies running on some of the largest supercomputers in the world, which will benefit from numerous applications high visibility scientific instruments. Additionally, the widespread adoption of ZFP in industry and academia will help advance many large-scale data management technologies, including HDF5, ADIOS, OpenZGY, OpenVisus, and Zarr.
The development of a high-performance SYCL door from ZFP on acceleration architectures supporting multiple vendors will benefit several highly visible supercomputing applications and better demonstrate the power of an open, standards-based software ecosystem.
“The work of the University of Utah and Lawrence Livermore National Laboratory in developing a high-performance SYCL-based ZFP library helps the availability of large-scale scientific data for high-performance computing architectures, enabling exascale applications to address multiple architectures. of accelerators, ”said Scott Apeland, senior director or Ecosystem Programs for Intel developers. “This latest Center of Excellence will show how open, standards-based development of an API 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 investigations and innovations, followed by practical validation and distribution to wider communities. CEDMAV’s previous collaborations with LLNL include shared research projects, dual-appointed staff, trainees and postdocs.
“It is an honor for CEDMAV to establish this oneAPI Center of Excellence in collaboration with LLNL. This will give a great opportunity to consolidate our collaboration and expand it with the support and collaboration of Intel engineers, “said Valerio Pascucci, founding director of CEDMAV and former leader of the CASC Data Analysis group at LLNL.” It is exciting to see the emergence of the oneAPI programming model that we intend to fully embrace in this project. In particular, the SYCL cross-platform abstraction will enormously increase the productivity of our teams in creating high-performance code that works efficiently on modern and heterogeneous architectures. Different hardware-software architectures are becoming ubiquitous in high-performance systems and oneAPI technology will greatly increase ZFP’s impact across a broad spectrum of applications. “
Information on CEDMAV
The University of Utah CEDMAV is internationally recognized for its activities involving theoretical and algorithmic research, systems development, and implementation of extreme data management tools. This research lies at the intersection of scientific visualization, big data management, HPC, and data analytics.
The Center for Applied Scientific Computing serves as LLNL’s window into the broader research communities of computer science, computational physics, applied mathematics, and data science. With partners from academia, industry and other government laboratories, it conducts world-class scientific research and development on issues critical to national security.
Information about an API
oneAPI is an open, unified, multi-architecture programming model for CPU and acceleration architectures (GPU, FPGA and others). Standards-based, the programming model simplifies software development and offers uncompromising performance for accelerated computation without proprietary lock-in, while allowing integration of existing code. With oneAPI, developers can choose the best architecture for the specific problem they are trying to solve without having to rewrite the software for the next architecture and platform.