The University of Tennessee, Knoxville, announced today the creation of a new Intel oneAPI Center of Excellence to provide solutions in high performance computing (HPC) and visualization using oneAPI. The center will focus on two projects: porting the open-source HPC Ginkgo library to oneAPI for cross-architecture support, and expanding its Intel Graphics and Visualization Institute of XeLLENCE to enable high-end visualization as a service through oneAPI.
oneAPI is an open sourced, unified programming model that simplifies development across multiple types of architectures (CPUs, GPU, FPGAs, and other accelerators). It’s based on industry standards and delivers uncompromised performance for accelerated compute, and removes constraints of proprietary programming models.
Two UT researchers have active projects at the center: Hartwig Anzt, a research scientist in UT’s Innovative Computing Lab, and Jian Huang, a professor in the Tickle College of Engineering’s Min H. Kao Department of Electrical Engineering and Computer Science.
Together, the two projects will boost computational science by opening the path toward improved multiarchitecture systems and software compatibility, while putting UT at the forefront of the latest innovations in the field. This work including technology development, optimization, validation, research, and curriculum design will also help increase development productivity and ecosystem adoption of oneAPI.
Translating Ginkgo to oneAPI
Anzt’s project is focused on porting the Ginkgo library to oneAPI. HPC Ginkgo is a modern C++ math library—a common programming language—for scientific HPC. The center will develop high-performance linear algebra components to achieve close-to-peak performance on multiple hardware architectures, which are then integrated and adopted by HPC projects to accelerate computing. This work extends Ginkgo library support to other accelerators—including taking advantage of current and future Intel® Xe GPUs.
oneAPI programming helps solve current development challenges where, typically, each architecture requires a different coding language and tool chain, forcing developers to rewrite programs and applications for different platforms. By using oneAPI’s standards-based Data Parallel C++ (DPC++), an implementation of the SYCL cross-architecture specification, developers can create a single source code where sparse linear algebra code can seamlessly run on multiple types of architectures.
This helps streamline software development for HPC initiatives, including research conducted by UT faculty and students, and encourages collaboration across research areas.
“Intel is an internationally recognized leader in the computing industry, and the establishment of this center is a great indication of the global impact of the tremendous HPC work being done here, with 35 ongoing projects in numerical linear algebra, performance evaluation and benchmarking, and distributed computing just in the ICL alone,” said Tickle College of Engineering Dean Matthew Mench, the Wayne T. Davis Dean’s Chair in the college. “We’re excited about this opportunity and eager to see the innovations that it helps develop.”
“The University of Tennessee’s work porting the Ginkgo project to oneAPI expands the choice of deployment hardware and providing support for Intel Xe architecture will deliver unprecedented performance to accelerate HPC’s journey to exascale,” said Jeff McVeigh, Intel vice president and general manager of the Super Compute Group.
Anzt, consultant at the ICL, will lead the oneAPI Center of Excellence with a team of experienced sparse linear algebra researchers. The center will conduct research and use oneAPI technology to contribute to the open specification and advance ecosystem adoption. The center’s work will help prepare Ginkgo for the Aurora supercomputer at Argonne National Laboratory.
Aurora is predicted to be one of the first exascale computers in the US, with applications across a multitude of fields, including clean energy production, nuclear safety, and cancer research.
Additionally, Anzt participates on the oneAPI Math Kernel Library (oneMKL) Technical Advisory Board to provide Sparse Linear Algebra expertise and feedback on the oneAPI libraries, compilers, and tools to help support researchers and developers around the world to take advantage of the resulting improvements. Further leveraging the university’s expertise, oneAPI curriculum modules will be developed and included in UT’s coursework and open-sourced to bring oneAPI programming skills to students worldwide in 2022.
Transforming Graphic Visualization
Huang’s project expanding UT’s Intel® Graphics and Visualization Institute of XeLLENCE (Intel® GVI) as an Intel® oneAPI Center of Excellence will provide high-end visualization as a service that can be instantly available and universally accessible. This transforms the paradigm that high-end visualization is heavy-weight, cumbersome to deploy, and expensive to use.
This service will use Intel® oneAPI technologies such as ray tracing/rendering and data and media processing to accelerate computing across heterogeneous architectures (CPUs, GPUs—including upcoming Intel® Xe architecture, and other accelerators). In an agile manner, developer communities can then compose visual dashboards that disseminate knowledge and insights broadly with interactive high-end rendering as the core.
To bring this vision to life, Professor Huang will create specialized visualization and rendering services, known as Visualization Cloud Instances (VCI). These VCIs will be containerized and easily deployed on HPC computers, on-prem institutional clusters, as well as through cloud service providers (CSPs). This multiple location strategy provides flexibility for app developers and system administrators with an unprecedented amount of freedom to choose software architectures that are optimal for their users.
VCI will be scalable, and cross-architecture-ready for CPU platforms with or without GPUs or other accelerators to maximize compute.
The university will also develop a curriculum on how interactivity enhances educational outcomes, and how to create and modify VCI instances and novel functionalities using Intel® oneAPI Toolkits. This training will be delivered to its students and later made public to the science and education communities.
“The university is delighted to work with Intel in developing this open innovation platform,” said Deborah Crawford, UT’s vice chancellor for research. “We know it will unleash the development of a wide range of applications, products, and services that will simultaneously advance science, innovation, and discovery, while also enhancing the health, wealth, and prosperity of communities here in Tennessee and around the world.”
“Professor Huang’s extensive background in technical computing visualization with an incredible ability to shed light on scientific simulations with significant complexity and results led him to spearhead creating powerful visualizations more easily with web-based tools like Tapestry. I look forward to our continued collaborations democratizing scientific visualization not only for scientists, but for everyone via cloud computing. By utilizing oneAPI heterogeneous programming and single source code using large-scale capable development tools, Huang democratizes scientific visualization code development with ease of use for academia and industry,” said Jim Jeffers, Senior Principal Engineer and senior director of Intel Advanced Rendering and Visualization.
About oneAPI
oneAPI is an open, unified and cross-architecture programming model for CPUs and accelerator architectures (GPUs, FPGAs, and others). Based on standards, the programming model simplifies software development and delivers uncompromised performance for accelerated compute without proprietary lock-in, while enabling the integration of legacy code. With oneAPI, developers can choose the best architecture for the specific problem they are trying to solve without needing to rewrite software for the next architecture and platform.
About Intel® oneAPI Rendering Toolkit
The Intel® oneAPI Rendering Toolkit is a set of open source libraries that enables creation of high-performance, high-fidelity, extensible, and cost-effective visualization applications and solutions. Its libraries provide rendering kernels and middleware for Intel® platforms for maximum flexibility, performance, and technical transparency. The toolkit supports Intel CPUs and future Xe architectures (GPUs). It includes the award-winning Intel® Embree, Intel® Open Image Denoise, Intel® Volume Kernel Library, Intel® OSPRay, Intel® OpenSWR, and other components and utilities.
© Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries.