Intel® Academic Program for oneAPI
oneAPI Centers of Excellence
The oneAPI centers deliver open-source code ports, additional hardware support, new technologies and services, and curriculum to increase oneAPI ecosystem adoption.
Argonne National Laboratory
Kalyan Kumaran's team collaborates with oneAPI groups to improve support for and the performance of various components critical to a range of science workloads for high-performance computing (HPC), data, and AI. The team's efforts—including the preparation of performant compilers, libraries, and AI frameworks for next-generation GPUs, as well as hardware testing, and application and profiling-tool development—will enhance the science productivity of these workloads on Aurora, the first exascale supercomputer of Argonne National Laboratory.
Charles University
Alexander Wilkie is focused on using the Intel® oneAPI Rendering Toolkit to help innovate sky model lighting simulations and integrate 3D and AI technologies for multiple usages. The center will also extend its clear sky model and other visualization technologies to oneAPI to support accelerator architectures, including upcoming Xe Architecture from Intel.
Chinese Academy of Sciences
Guangming Tan conducts research on parallel programming and algorithms, domain-specific architecture, and bioinformatics. Based on the open oneAPI specification, his team will extend the oneAPI unified programming framework to support China's local accelerators and use oneAPI to develop full-stack, open-source software. He has more than 70 papers published in his research field in international journals and conferences such as IEEE Transactions.
Heidelberg University / Heidelberg University Computing Center (URZ)
Askel Alpay and Dr. Vincent Heuveline are focused on adding advanced DPC++ capabilities to hipSYCL to further cross-architecture computing.
Lobachevsky State University of Nizhni Novgorod
Iosif Meerov and Arkady Gonoskov are focused on porting the High-Intensity Collisions and Interactions (Hi-Chi) application to oneAPI to speed research on complex classical and quantum systems.
Northern Illinois University
Michael Papka and Joe Insley are focused on using oneAPI to accelerate and simplify complex data through visual transformations. The center will integrate oneAPI programming and the Intel® oneAPI Rendering Toolkit into its toolset to enable students’ research and sharpen their skills for writing single-source heterogeneous code to target a variety of architectures (CPUs and accelerators). Professors will also incorporate oneAPI into their courses.
Oak Ridge National Laboratory
Principle investigator Dr. Jin Zheming is a research scientist and an expert on high-level synthesis and heterogeneous parallel computing. His research will focus on direct CUDA* porting.
Old Dominion University
Mohammad Zubair (with NASA* support) optimizes unstructured-grid computational fluid dynamics (CFD) kernels and develops efficient implementations on Intel® CPUs and GPUs with Xe Architecture to help solve some of the nation’s most complex aerodynamic problems.
Stockholm University / KTH Royal Institute of Technology
Erik Lindahl is focused on porting the GROMACS molecular dynamics package over to oneAPI and sharing his experience with the broader ecosystem.
University of California at Berkeley
Kurt Keutzer is focusing on producing energy-efficient algorithms and implementations for deep learning’s most computationally intensive workloads. The algorithms will be used in training recommendation systems and natural-language understanding systems to help address important AI challenges. The center will use the oneAPI Deep Neural Network Library (oneDNN) and the oneAPI Collective Communications Library (oneCCL) to optimize this work, and oneAPI programming will significantly ease the development of portable implementations across multiple types of architectures.
University of California at Davis
Kwan-Liu Ma is focused on using oneAPI and the Intel oneAPI Rendering Toolkit to deliver high-performance, high-fidelity data visualization and analysis applications that scale across Intel® CPUs and GPUs, including Xe Architecture from Intel.
University of Cambridge - Stephen Hawking Centre of Theoretical Cosmology
Paul Shellard is focused on advancing cosmological research, open-source code development, and in-situ compute and visualization, as well as teaching computational and visualization coding techniques. Through using the Intel® oneAPI DPC++/C++ Compiler and Intel oneAPI Rendering Toolkit, the Centre will optimize large-scale cosmology workflows that help drive new discoveries.
University of Durham
Tobias Weinzierl conducts research on task-based programming using oneAPI development and championing oneAPI training. The center will extend the hyperbolic PDE engine ExaHyPE into a OneExaHyPE code that scales across a wide variety of GPU-accelerated machines. The algorithmic and methodological insights will be beneficial for many other simulation codes. For training, the center will organize oneAPI workshops and tutorials that will be open to Durham students and faculty, and available to colleagues in the UK and beyond.
University of Illinois at Urbana-Champaign
Emad Tajkhorshid and David Hardy are focused on using the heterogeneous programming model to accelerate computing for biomolecular research using NAMD.
University of Stuttgart (VISUS)
Guido Reina is focused on cross-architecture in-situ visualization across CPUs, GPUs, and other accelerators. He will use oneAPI programming and the Intel oneAPI Rendering Toolkit for advanced ray tracing and parallelization capabilities to evolve the scalable, in-situ capable back end of the visualization framework MegaMol, which helps study and understand physical processes in fields like thermodynamics and porous media research.
University of Tennessee at Knoxville
Jian Huang and Hartwig Anzt are focused on oneAPI adoption through HPC and high-end visualization projects supporting heterogeneous architectures, research, and curriculum. For HPC, Professor Antz and his team will port the open-source HPC Ginkgo library to oneAPI to take advantage of the current and future GPUs built on Xe Architecture from Intel.
For visualization, Professor Huang’s team will provide high-end visualization as a service that can be instantly available and universally accessible. They will use oneAPI technologies such as advanced ray tracing or rendering, and data and media processing to accelerate computing across multiarchitectures.
University of Texas – Texas Advanced Computing Center (TACC)
Paul Navratil is focused on powering extreme-scale remote visual analysis using oneAPI and the Intel oneAPI Rendering Toolkit. TACC will use oneAPI single source programming, Intel®-optimized ray tracing libraries, plus Intel CPU and Xe Architecture GPU hardware for scientific simulations that employ algorithms and techniques to deliver new science through fast in-situ visual analysis. This will enable flexible, high-fidelity, high-performance interactive analysis across TACC computing platforms without having to implement device-specific routines for each system.
University of Utah - Scientific Computing and Imaging (SCI) oneAPI Center
Chris Johnson is focused on improving large-scale simulations, data analytics and visualization for scientific workflows. To do this, he will advance research and development, teach the latest visual computing innovations in ray tracing and rendering, and use oneAPI to deliver high-performance compute across heterogeneous architectures (CPUs, GPUs including the upcoming Xe architecture from Intel, and other accelerators).
Academic Projects
Efficiency and Productivity for Decision-Making on Low-Power CPU and GPU SoCs
The University of Málaga is researching how to solve the large-scale Markov Decision Process efficiently on low-power computing platforms using 12 different parallel code implementations. The project highlights the use of oneAPI libraries and languages, such as Intel® oneAPI Threading Building Blocks and DPC++.
TAU Parallel Performance System
The TAU Parallel Performance System is a profiling and tracing tool framework for performance analysis of parallel programs. It has been integrated to the runtime of three different GPU vendors to profile a SYCL application to emphasize the benefits of an open industry standard.
oneOligo
OneOligo is an EU-funded project that implements synthetic DNA as storage to meet the demands of a data-driven economy as data generation rate outpaces the storage density of currently available media. Synthetic DNA possesses key properties that make it relevant for archival storage density, ruggedness, and it is easy and cheap to replicate like organic DNA.
Open-Source Scientific Applications and Benchmarks
This repository contains a collection of data-parallel programs migrated from CUDA. They compare the performance of the SYCL, DPCT-generated, and OpenMP implementations of each program. This is an example of how SYCL as an open standard can increase productivity and, in time, facilitate the adoption of SYCL, migrating applications for multiplatform support.
Professor Insights
"It’s exceptionally important to us that GROMACS is able to make efficient use of all the fastest supercomputers in the world, and the upcoming exascale machines powered by Intel hardware will make it possible to simulate processes we could not even imagine a few years ago, such as how a virus binds to proteins on a cell and infects it."
— Dr. Erik Lindahl, professor of biophysics, GROMACS Development Team, KTH Royal Institute of Technology
"By using Intel® DPC++ Compatibility Tool, hand-tuned CUDA codes were migrated with almost no effort (only a few small fixes applied, with no prior knowledge on DPC++). With migrated code, we obtained speed-ups of up to 9.5x on DG1 (versus Gen 9.5 GPU) for two-way and three-way epistasis detection on different datasets."
— Professor Aleksandar Ilic, INESC-ID, Universidade de Lisboa, Portugal
"My area of interest is computer architecture, more specifically FPGAs. Hardware design based on high-level specifications has evolved considerably and is likely to make the use of HDL unnecessary in most projects. SYCL seems to me a promising standard, mainly because it is open."
— Dr. Ricardo Menotti, Federal University of São Carlo
"One of our strategic goals is to make a measurable contribution to the transfer of new technologies from research to industrial application, and of course, to continuously expand our expertise and research efforts in the field of supercomputing. The oneAPI CoE will allow us to do both."
— Dr. Vincent Heuveline, professor and director of the Computing Center of Heidelberg University
Benefits
Universities and students who participate in the academic program will gain access to Intel® DevCloud hardware accelerators and compute nodes, Intel® DevMesh to showcase work, and technical support.
Intel® DevCloud
Develop, run, and optimize your Intel® oneAPI solution in the Intel DevCloud—a free development sandbox to learn about and program oneAPI cross-architecture applications. Get full access to the latest Intel® CPUs, GPUs, and FPGAs, Intel oneAPI toolkits, and the new programming language, DPC++.
Training
Essentials of Data Parallel DPC++
Learn the fundamentals with this seven-part, self-paced course which includes hands-on practice and sample code.
Intel® DevMesh
This is a community portal for students, professors, and developers to share their work and collaborate while building a professional profile of amazing work and activities.
See how Centers of Excellence (CoE), universities, students, companies, and developers innovate using Intel oneAPI toolkits and program resources.
Intel® oneAPI Toolkits
These toolkits include optimized tools, languages, and libraries for building and deploying high-performance, data-centric applications and solutions across CPUs, GPUs, FPGAs, and other accelerators.
Events
Webinars
Get access to the latest information and training about oneAPI through live and on-demand webinars.
Meetups & Workshops
Participate in virtual meetups with Intel oneAPI Toolkit experts and your peers in the developer and academic communities.