CRPL -- Computational Research Programming Lab

Example Project

Tue, 23 Jun 2020 12:33:45 +1000

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Donec accumsan dapibus purus, vel maximus erat facilisis vel. Suspendisse eu ipsum consequat, sodales nulla vel, faucibus velit. Fusce imperdiet lacus nec cursus auctor. In vulputate arcu et nulla placerat, nec pulvinar quam rhoncus. Praesent vel placerat nunc. Mauris elit felis, pulvinar egestas risus ut, fermentum ullamcorper orci. Integer tempor, neque sed fermentum eleifend, ligula est tincidunt purus, sodales semper elit enim ut mi. Vestibulum commodo tristique hendrerit. Duis interdum tristique augue id aliquam. Sed et tellus aliquam, accumsan erat at, facilisis tortor. Donec ac sapien sed lectus ullamcorper ultrices. Nullam placerat mauris nec urna rhoncus, vitae faucibus nulla venenatis. Donec vestibulum justo auctor, vestibulum diam quis, egestas erat. Vivamus volutpat et metus ut fermentum. Donec volutpat neque nec quam lacinia, nec faucibus felis ullamcorper.

Capabilities implementation and testing on normal and tumor tissue cells

Tue, 26 May 2020 11:04:20 +0200

NCI-DOE Drug response modeling

PI: Sunita Chandrasekaran

Students: Vineeth Gutta

Collaborators: National Cancer Institute at Frederick National Lab: Sean Black, Eric Stahlberg, Satish Ranganathan & Matthew Beyers

Funding Agency and Period : National Cancer Institute, National Institute of Health, Leidos Biomedical Research and 11/21 - Present

Project Duration: 11/21 - 01/25

Project Summary:

ECP-CANDLE (Cancer Distributed Learning Environment) is a collection of models and libraries that include 3 Pilots where each tackles a separate problem. Pilot 1 focuses on the drug response problem, pilot 2 on RAS pathway problem, and pilot 3 on high level treatment strategy problem. This work only involved pilot 1. The fundamental challenge drug response modeling faces is the lack of data for drug response on patients because human trials given the number of possible drugs makes it impossible to test them all. Therefore drug response models are trained on cancer cell-lines (grown in the lab) and these models do not perform well on patient tissue data. Some example models include single drug response model and Combo model with the difference being the latter using a combination of drugs to predict growth of cancer cell line. Both use multilayer perceptron with convolution like layers for the model architecture. The current work focuses on modeling drug response for gene expression data from CCLE (Cancer cell line encyclopedia) and drug sensitivity data from GDSC (Genomics of drug sensitivity in cancer). Existing CNN models from Single drug response and Combo models performed worse on the CCLE/GDSC data compared to the original NCI60 datasets. So we needed to improve the model itself. The XGBoost model improved the performance for CCLE/GDSC data when evaluated using R squared error metric. For future work we are exploring the use of Autoencoders to encode features such as drug descriptors or gene expressions to improve the performance of CNN models.

Mon, 01 Jan 0001 00:00:00 +0000

DOE S4PST, Next Generation Science Software Technologies Project

Wed, 17 Aug 2022 11:04:20 +0200

Co-PI: Sunita Chandrasekaran

Current Students: Aaron Jarmusch, Giorgi Gvalia

Collaborators: Oak Ridge National Laboratory, Lawrence Livermore National Laboratory, Sandia National Laboratory among others

Funding Agency and Period: Department of Energy, 01/24 - 12/28

Project Duration: 01/24 - Present

Project Summary:

Our goal is to enable a more productive ecosystem for the sustainability of high-performance computing (HPC) software for node level programming systems and tools (PST). The scope includes exploratory work to identify technical, economic and social requirements important to the HPC community and to plan for a sustainable strategy to deliver robust, trustworthy software in a thriving community invested in addressing the challenges of the current and future heterogeneous computing landscape.

Mon, 01 Jan 0001 00:00:00 +0000

ORNL CAAR: Preparing PIConGPU, a Plasma Physics application for Frontier Exascale system

Tue, 26 May 2020 11:04:20 +0200

PI: Sunita Chandrasekaran

UDEL Students: Vineeth Gutta

Collaborators from HZDR/CASUS: Michael Bussmann, Alexander Debus, Guido Juckeland, Rene Widera, Sergei Bastrakov, Thomas Kluge, Klaus Steiniger, Jeff Kelling, Richard Pausch, Marco Garten, Felix Meyer, Matthias Werner, Franz Poeschel, Jeyhun Rustamov

Funding Agency and Period: CASUS – Center for Advanced Systems Understanding, Helmholtz-Zentrum Dresden-Rossendorf e.V. (HZDR) Untermarkt 20, D-02826 Görlitz, Germany, May 2020 to March 2022

Project Duration: 10/01/2019 - present

Project Summary:

Mon, 01 Jan 0001 00:00:00 +0000

XBLang: An Extensible Compiler Infrastructure Based on MLIR

Tue, 26 May 2020 11:04:20 +0200

PI: Sunita Chandrasekaran

Students: Fabian Mora

Collaborators: UDEL: Felipe Cabarcas; Lawrence Livermore National Laboratory: Johannes Doerfert

Funding Agency and Period: National Science Foundation (NSF) 1814609 and 10/01/2018 – 09/30/2023

Project Duration: 08/01/2019 - 12/24

Project Summary:

The goal of this research project is to develop a compiler infrastructure that facilitates research on programming abstractions, compiler optimizations, parallel programming models, and compiler construction.

For this purpose, recently, we have designed XBLang, a compiler infrastructure developed on top of the widely-used MLIR compiler infrastructure. XBLang uses a novel approach where the traditional Abstract Syntax Tree (AST) used in compiler frontends is replaced by an MLIR representation. This approach proves extremely powerful as it allows the use of all the machinery available in MLIR to analyze, optimize, and introduce new programming abstractions.

Mon, 01 Jan 0001 00:00:00 +0000

LLM4VV

Tue, 26 May 2020 11:04:20 +0200

LLM4VV: Developing LLM-Driven Testsuite for Compiler Validation

PI: Sunita Chandrasekaran

Students: Zack Sollenberger, Saieda Ali Zada, Rahul Patel, Aaron Jarmusch

Past Students: Christian Munley

Funding Agency and Period: OpenACC and 07/2017 – present

Project Duration: 04/30/23 – present

Project Summary:

Large language models (LLMs) are a new and powerful tool for a wide span of applications involving natural language and demonstrate impressive code generation abilities. The goal of this work is to automatically generate tests and use these tests to validate and verify compiler implementations of a directive-based parallel programming paradigm, OpenACC. To do so, in this paper, we explore the capabilities of state-of-the-art LLMs, including open-source LLMs – Meta Codellama, Phind fine-tuned version of Codellama, Deepseek Deepseek Coder and closed-source LLMs – OpenAI GPT-3.5-Turbo and GPT-4-Turbo. We further fine-tuned the open-source LLMs and GPT-3.5-Turbo using our own testsuite dataset along with using the OpenACC specification. We also explored these LLMs using various prompt engineering techniques that include code template, template with retrieval-augmented generation (RAG), one-shot example, one-shot with RAG, expressive prompt with code template and RAG. This paper highlights our findings from over 5000 tests generated via all the above mentioned methods. Our contributions include: (a) exploring the capabilities of the latest and relevant LLMs for code generation, (b) investigating fine-tuning and prompt methods, and (c) analyzing the outcome of LLMs generated tests including manually analysis of representative set of tests. We found the LLM Deepseek-Coder-33b-Instruct produced the most passing tests followed by GPT-4-Turbo.

Mon, 01 Jan 0001 00:00:00 +0000

OpenMP VV: A Testsuite to Validate and Verify OpenMP Features which is apart of the ECP SOLLVE project

Wed, 17 Aug 2022 11:04:20 +0200

PI: Sunita Chandrasekaran

Current Students: Andrew Kalla

Past Students: Thomas Huber, Jose Manuel Monsalve Diaz, Joshua Davis, Nolan Baker, Kristina Holsapple, Jaydon Reap, Michael Carr, Nikhil Rao, Ray Escobar

Oak Ridge National Lab (OLCF): Swaroop Pophale, Seyong Lee, David Bernholdt

Funding Agency and Period: Department of Energy (Exascale Computing Project - SOLLVE) and 07/19 - 12/23

Project Duration: 07/19 – present

OpenMP V&V Site: https://crpl.cis.udel.edu/ompvv/

Mon, 01 Jan 0001 00:00:00 +0000

Validation and Verification Testsuite OpenACC

Wed, 17 Aug 2022 11:04:20 +0200

PI: Sunita Chandrasekaran

Current Students Nathan Graddon, Ryan Padrone, Aaron Jarmusch

Past Students: Christian Munley, Kyle Friedline, Aaron Liu, Vaidhyanathan Ravichandran, Daniel Horta, Will Gunter

Collaborators: NVIDIA, HPE, GCC and ORNL

Funding Agency and Period: OpenACC and 07/2017 - present

Project Duration: 09/01/2016 – present

OpenACC V&V Site: https://crpl.cis.udel.edu/oaccvv/

Project Summary:

The goal of this project is to build a validation and a verification testsuite for the OpenACC Programming Model. Since 2016, we have created 1081 functional/unit test codes to verify and validate implementations of OpenACC and its conformance to the OpenACC specification. Our project also includes an infrastructure that makes it easier to capture the tests that failed to compile, erred at runtime and which segments of a given test failed.

Mon, 01 Jan 0001 00:00:00 +0000

ACM SIGHPC Emerging Women Leader in Technical Computing Awarded to Dr. Sunita Chandrasekaran @ SC25

Thu, 20 Nov 2025 00:00:00 +0000

The ACM SIGHPC Emerging Woman Leader in Technical Computing (EWL/TC) Award is a prestigious international honor that recognizes women who are shaping the future of high-performance and technical computing. Established by the ACM Special Interest Group on High Performance Computing (SIGHPC), the award highlights women who have been actively engaged in HPC and technical computing research, education, and/or practice for roughly 5–15 years after their highest degree.

The award is presented at the annual SC (Supercomputing) conference, the flagship global conference for high-performance computing, networking, storage, and analysis. Beginning in 2022, EWL/TC became an annual award, with the recipient honored on stage during SC and recognized alongside other major ACM and IEEE awards.

SC 25

Mon, 17 Nov 2025 00:00:00 +0000

The Computational Research and Programming Lab (CRPL) returned from SC25 in St. Louis after an exciting week immersed in the world’s premier conference for high-performance computing. SC brings together researchers, students, industry partners, and national labs from around the globe for technical talks, tutorials, posters, Birds of a Feather (BoF) sessions, and a massive exhibit floor highlighting the latest in supercomputing and AI. Our group has built up deep experience attending SC over the years, and this year we were proud to bring a strong team of students and researchers who actively contributed across multiple parts of the program.

Aaron @ Argonne Training Program on Extreme-Scale Computing (ATPESC) 2025

Thu, 01 May 2025 00:00:00 +0000

Aaron just finished up an incredible two weeks at the Argonne Training Program on Extreme-Scale Computing (ATPESC) 2025.

A U.S. Department of Energy (DOE) sponsored program to develop key skills to design, implement and execute computational science and engineering applications on current and future high performance computing systems. Topics discussed included: numerical algorithms, performance modeling, GPU architectures, and scalable software tools.

Besides the technical content, one of the most valuable parts for Aaron was connecting with fellow researchers, developers, and industry leaders pushing the field of high-performance computing. His discussions with Kathy Yelick, Tim Mattson, and Jack Dongarra, made ATPESC a packed and eye-opening experience.

ECP SOLLVE: Scaling OpenMP With LLVM for Exascale Performance and Portability

Sun, 20 Apr 2025 00:00:00 +0000

PI: Sunita Chandrasekaran

UDEL Students and Software Engineers: Aaron Jarmusch, Aaron Liu, Felipe Cabarcas

Funding Agency: Department of Energy

Duration: 10/01/2020 – 12/31/2023

Collaborating organizations involved in SOLLVE include:
-Brookhaven National Laboratory (BNL)

-University of Delaware

-Argonne National Laboratory

-Lawrence Livermore National Laboratory

-Oak Ridge National Laboratory

-Rice University

-University of Illinois at Urbana-Champaign

These institutions collaborated to enhance OpenMP’s capabilities, ensuring that scientific applications can efficiently utilize the computational power of forthcoming exascale supercomputers like Frontier, Aurora, and Perlmutter.

Saieda's Summer Research - Chameleon Concierge: Retrieval-Augmented Generation (RAG) To Enhance Open Testbed Documentation

Tue, 01 Apr 2025 00:00:00 +0000

Chameleon Concierge uses Retrieval-Augmented Generation to help researchers navigate complex open science testbed documentation. When designing experiments, researchers struggle to find technical solutions across multiple documentation sources, often leading to support tickets or project abandonment. Our custom LLM search service generates accurate, cited responses to natural language queries. Results show RAG models perform comparably to proprietary LLMs while bridging the gap between researcher knowledge and static documentation, improving research productivity and infrastructure operations.

Zahra's Summer Internship: Improving Usability and Performance in cc-snapshot

Tue, 01 Apr 2025 00:00:00 +0000

During Zahra’s 2025 summer internship she had the incredible opportunity to collaborate with mentor Paul Marshall (UChicago) on the cc-snapshot project. This included:

Enhancing Usability and Performance in Experimental Environments Management

Maintaining an identical setup and reproducing environments is a challenge in high-performance computing (HPC) and research. HPC experiments are resource-intensive and depend on complex software environments. Existing methods, such as orchestration with containers, create controlled environments, but require careful setup and maintenance. However, snapshotting captures the complete state of a system in a single step, allowing researchers to automatically rebuild and restore identical environments. Although, concerns remain about snapshot efficiency and usability.

Rahul's ISACA Philadelphia Chapter Scholarship

Sat, 01 Mar 2025 00:00:00 +0000

The ISACA Philadelphia Chapter Scholarship is awarded to five students in the Greater Philadelphia area who are pursuing an associate, bachelor’s, or master’s degree in fields related to Information Systems, Information Technology, Cybersecurity, Computer Science, IT Audit, IT Compliance, or IT Governance.

CGO 2025

Sat, 01 Feb 2025 00:00:00 +0000

Saieda Ali Zada, Zahra Temori and Zachariah Sollenberger participated in the FastCode Challenge program co-located with CGO/PPoPP/HPCA/CC 2025 conference.

Saieda, Zahra, and Zachariah attended CGO/PPoPP/HPCA/CC 2025 conference as student volunteers. We attended sessions on programming language, compilers and LLMs and met researchers working in this area.

IPDPS Article

Sat, 01 Feb 2025 00:00:00 +0000

Nominated for the Best Paper Award @ IPDPS 2025, Milan, Italy.

Increasing HPC cluster sizes and large-scale simulations that produce petabytes of data per run, create massive IO and storage challenges for analysis. Deep learning-based techniques, in particular, make use of these amounts of domain data to extract patterns that help build scientific understanding. Here, we demonstrate a streaming workflow in which simulation data is streamed directly to a machine-learning (ML) framework, circumventing the file system bottleneck. Data is transformed in transit, asynchronously to the simulation and the training of the model. With the presented workflow, data operations can be performed in common and easy-to-use programming languages, freeing the application user from adapting the application output routines. As a proof-of-concept we consider a GPU accelerated particle-in-cell (PIConGPU) simulation of the Kelvin- Helmholtz instability (KHI). We employ experience replay to avoid catastrophic forgetting in learning from this non-steady process in a continual manner. We detail challenges addressed while porting and scaling to Frontier exascale system.

Wed, 01 Jan 2025 00:00:00 +0000

The opportunity to work under Prof. Sunita Chandrasekaran during my PhD proved to be invaluable for my research, collaborations, and career. I will always be grateful for her mentorship, guidance, and expertise during my time here. She actively fosters a collaborative and enriching environment in the lab. It also helps that everyone here is wonderful, smart, and fun to be around :).

The various areas of research I had the chance to explore during my PhD and the skills I developed during my time here are proving to be useful in my current role and I’m confident they will continue to do so for the rest of my career. My work here gave me access to some of the worlds leading supercomputers and thus allowed me the ability to work on problems I could not have tackled anywhere else. Sunita always encouraged me to become a well rounded researcher through effective communication and helped to find the most impactful venues to practice these skills. The collaborative research I engaged in beyond the lab allowed me to leverage the expertise of those outside the lab to augment my work done here.

Sun, 01 Dec 2024 00:00:00 +0000

I got my PhD in Computer Science at the University of Delaware. From the start of my PhD, I worked as a Research Assistant at the CRPL lab under the supervision of Prof. Sunita Chandrasekaran. She is a fantastic professor, advisor, and supervisor, always listening, providing guidance, and pushing people to improve. Further, the environment she fosters at CRPL creates a dynamic of collaboration, innovation, and respect, all essential characteristics to make graduate students triumph in the journey through grad school.

Year-end Newsletter 2024

Sun, 01 Dec 2024 00:00:00 +0000

Here is a newsletter with our group’s activities and publications for the calendar year 2024!

SC 24

Sun, 17 Nov 2024 00:00:00 +0000

Back from SuperComputing 24: Highlights from the University of Delaware’s HPC Journey

Late November 2024 our team returned from Atlanta, Georgia, where we participated in SC24, the world’s leading conference for high-performance computing. It was a privilege to reconnect with the HPC community and showcase our latest achievements. Here are the highlights:

The University of Delaware maintained a strong presence at SC24 with a dedicated booth, highlighting cutting-edge research and collaborations. This year marked a significant milestone, with a record-breaking conference attendance of approximately 18,000, and we were excited to demonstrate how our work is shaping the future of HPC. Our booth featured a 3D printer giveaway alongside other swags, making our participation both informative and engaging.

Visit from HPE Researcher Dr. Utz-Uwe Haus

Tue, 16 Jul 2024 00:00:00 +0000

We are excited to announce that Utz-Uwe Haus, Head of the HPE HPC/AI EMEA Research Lab (ERL), visited our lab to give a talk on July 15th 2024.

About Utz-Uwe Haus

Utz-Uwe Haus leads Research and Development for HPE in Europe. He studied mathematics and computer science at the Technical University of Berlin (TU Berlin) and holds a doctorate in mathematics from the University of Magdeburg (Germany). Haus has an extensive background in nonstandard applications of mathematical optimization in fields such as chemical engineering, material science, and systems biology.

Mid-Year Team Update 2024

Thu, 11 Jul 2024 00:00:00 +0000

Exiting Changes and New Faces

As we hit the mid-year mark of 2024, our computer science lab has experienced a rollercoaster of changes, marked by both farewells and fresh beginnings. This update serves to highlight the most significant events and introduce the new talents joining our journey.

Farewell and New Beginnings

First, we bid a heartfelt farewell to our postdoctoral researcher, Felipe Cabarcas. Felipe has returned to the Universidad de Antioquia in Medellin, Colombia, to further his academic career. His contributions to our lab have been invaluable; from training new talents to bringing a cheerful and positive atmosphere, Felipe’s presence will be greatly missed. We wish him all the best in his future endeavors.

Wed, 01 May 2024 00:00:00 +0000

The Computational Research and Programming Lab (CRPL) at the University of Delaware is an awesome group. Not only are they leaders in the field of high-performance computing and AI, they also are a wonderful group of people, which is most important! Dr. Sunita Chandrasekaran leads various projects within the group and gives her students the opportunity to own research projects, collaborate with industry experts, and present their work in various conferences and venues. I joined CRPL as a high school student, and stayed throughout my undergraduate education. Along the way, we published multiple papers, developed open-source projects, won various awards and met many great researchers and professionals. My career in HPC and AI is largely thanks to Dr. Chandrasekaran and everyone in CRPL. To anyone who is interested in research in HPC or AI, CRPL is the place to be!

SC 23

Mon, 25 Sep 2023 00:00:00 +0000

We returned from our trip to Denver, Colorado, for SC23, one of the largest high-performance computing conferences in the world, in November. We were thrilled to have the opportunity to participate and attend this exciting event. Here are some highlights from SC23:

For the first time since 2019, the University of Delaware had a booth at SC23, and we were excited to showcase the remarkable work being done. This year marked a significant milestone for us, and we were so excited to share the groundbreaking developments and research with the community.
Our OpenACC V&V Testsuite team, including Christian Munley and Aaron Jarmusch, has been hard at work, and were eager to showcase their latest innovations and achievements at our booth. Stay tuned for updates on their cutting-edge projects that will be featured in 2024.
The ECP SOLLVE V&V project is making a significant impact at SC23, with contributions from Andrew Kallai, Felipe Cabarcas, and Eric Wright. They had both a research poster and a talk on their SIMD work, providing valuable insights and contributions to the conference.

UD's DS + AI Hackathon

Wed, 19 Jul 2023 00:00:00 +0000

Introduction

The University of Delaware’s Data Science Institute (DSI) and AI Center of Excellence (AICoE) recently hosted a dynamic 3-day hackathon, bringing together exceptional minds to tackle complex challenges in data science, artificial intelligence (AI), and high-performance computing (HPC). In a resounding testament to their problem-solving abilities and innovative spirit, two standout teams from the Computer Research and Programming Lab (CRPL) secured prestigious awards. The Kendal Transformers, led by Nikhil Rao and Mauricio Ferrato, clinched the Technical Complexity award, while the Medical Marauders, under the leadership of Andrew Kallai, emerged victorious with the Impact award.

Wed, 31 May 2023 00:00:00 +0000

I am extremely grateful for the time I spent under the mentorship of Prof. Sunita Chandrasekaran and the support I received from the other members of CRPL.

I first joined when I was a computer engineering undergraduate student and at the time I was not sure where I wanted to take my career. While in the lab I was able to partake in a vast range of research areas in accelerated computing. My journey at CRPL started from studying parallel programming models and high performance computing all the way to graduating with a PhD in Computer Science, applying machine learning and AI to real world scenarios. Sunita always encouraged the different avenues I took, and with the extensive knowledge from my collaborators and peers in the lab, it always felt like I was on the right path. I am now a Solutions Architect at NVIDIA, tackling some of the world’s hardest challenges, and thanks to the foundation that was created for me at CRPL I feel like I belong.

Mon, 01 May 2023 00:00:00 +0000

During my master’s program at the University of Delaware, I worked as a graduate research assistant at CRPL and I genuinely loved it. Under Sunita’s guidance, I didn’t just improve my technical and research skills, I also grew a lot personally and professionally. CRPL fosters a collaborative work environment by encouraging open discussions, creative problem solving, constructive criticism and mutual support among all. I’m really grateful for my time at CRPL. It prepared me for my career in software development and gave me a solid foundation of skills and relationships that I’ll cherish for a lifetime.

Best Poster Award at 2023 DARWIN Symposium

Thu, 23 Feb 2023 00:00:00 +0000

OpenACC funded OpenACC V&V (Verification and Validation) and DOE funded OpenMP V&V (part of the DOE Exascale Computing Project SOLLVE) testsuites are a framework of suites and tools aimed at ensuring the correctness and performance of OpenACC and OpenMP compiler implementaions. OpenACC and OpenMP are popular programming models for accelerating applications on heterogeneous computing systems. Our poster that presented both the suites was selected for the Best Poster Award at the recently held DARWIN symposium (Feb 23). We are so grateful for the recognition and honor as it highlights the significance and impact of our testsuite running on UDEL clusters like DARWIN on campus.

SC 22 Highlights

Sat, 03 Dec 2022 00:00:00 +0000

We recently returned from our trip to Dallas, TX to participate and attend SC22. SC22 is one of the largest high performance computing conferences in the world and it was amazing to be able to attend. Some highlights from the week are below.

Jaydon Reap was selected to join sixteen other students and help directly with SCiNet. The networking team behind SC. He worked with the fiber team and worked to deploy the backbone infrastructure for the conference network. Jaydon worked directly in one of the Distributed Network Operations Centers (DNOCs) and helped provide internet connectivity to booths in the exhibit floor.

PIConGPU running on ORNL's Frontier

Fri, 18 Nov 2022 00:00:00 +0000

Just how fast can the world’s fastest supercomputer go? Maybe even faster than imagined.

Researchers studying plasma physics for particle accelerators recently used the Oak Ridge Leadership Computing Facility’s Frontier supercomputer to achieve a speedup by as much as eightfold in their code’s performance – more than double the improvement expected. The achievement raises scientists’ already high hopes for the world’s first exascale supercomputing system.

For more

Big Data Analytics and Machine Learning

Wed, 02 Nov 2022 11:04:20 +0200

PI: Sunita Chandrasekaran

Students: Mauricio H. Ferrato, Mathias Heider

Collaborators: Nemours Children’s Hospital: Erin L. Crowgey, Karl Franke

Collaborators: University of Delaware: Adam Marsh

Funding Agency: Nemours Children’s Hospital

Duration: 09/01/2018 – 06/30/2023

Project Summary:

The goal of this project is to build predictive models for rare disease outcomes by using machine learning and deep learning approaches. During the long process of the project we have focused on a few diseases: Sickle Cell Disease (SCD), Parkinson’s Disease (PD), and Acute Myeloid Leukemia (AML). The goal of the SCD and PD projects was to predict adverse events, such as acute chest and vaso-occlusive crisis for SCD, at the individual level by utilizing genomics (variant data). The goal of the oncology project is to predict drug response as well as vital status using transcriptomics and ex vivo data. Machine learning has proven to be a potential approach for early detection and prevention of human diseases.

New Postdoc Felipe Cabarcas working with CRPL

Mon, 01 Aug 2022 00:00:00 +0000

We are excited to announce we have hired our lab’s first postdoc to help with various lab projects. Primarily working with the ECP SOLLVE project. We are excited to have Felipe join our team!

Felipe Cabarcas graduated in Electrical Engineering in 1999 from the Universidad Pontificia Bolivariana, Medellin, Colombia. He obtained a Master of Electrical Engineering in 2002 from the University of Delaware. And a Ph.D. in Computer Architecture from the Universidad Politécnica de Catalunya in 2010, working in the Barcelona Supercomputing Center. He has been professor of Electrical Engineering in the Universidad de Antioquia for more than 10 years. During the last 10 years he has been teaching digital design courses; and his research has focused mainly on Genomics, and Metagenomics.

Where were CRPL students this Summer?

Mon, 11 Jul 2022 00:00:00 +0000

Fabian Mora - NVIDIA
Mauricio Ferrato - NVIDIA
Vineeth Gutta - Pacific Northwest National Lab
Eric Wright - Argonne National Lab
Kristina Holsapple - MIT - Healthy ML Group of the Computer Science and Artificial Intelligence Laboratory (CSAIL)

January 2022 Highlights

Sat, 01 Jan 2022 00:00:00 +0000

Kristina Holsapple selected as Adobe Women-in-Technology Scholar 2022 scholar
Paper to CCGrid 2022 ACCEPTED!!!! Authors include former CRPL undergraduate researcher Miguel Zavala, who is now with AMD and Sunita Chandrasekaranand and several other academic members from SPEC HPG
Matthew Leinhauser, our PhD student’s paper on “Metrics and Design of an Instruction Roofline Model for AMD GPUs” accepted and published at ACM TOPC, 2022 !!!! Check out our publication
CRPL adds more members to the ECP SOLLVE project, Nikhil Rao, pursuing Masters in ECE, and junior undergraduate students Kristina Holsapple, Jaydon Reap and Michael Carr join

Mon, 31 May 2021 00:00:00 +0000

I worked with CRPL and Prof. Sunita Chandrasekaran for most of my four years of undergraduate education at the University of Delaware. In that time I had unique opportunities to work on multiple research projects, carry out internships at DoE national laboratories, co-author papers as an undergraduate, and get accepted to a competitive PhD program at a peer institution, all through CRPL and the mentorship Sunita provided. She is an exceptional mentor, going above and beyond for her students to make sure they are able to meet their goals, not to mention a visionary researcher and outstanding communicator and organizer for promoting high-performance computing education and research.

CRPL Virtual Webinar Series

Mon, 15 Mar 2021 00:00:00 +0000

This series included David Donofrio/John Leidel from Tactical Computing Labs, Florina Ciorba from UBasel, Ronnie Chatterjee from LBNL, Ian Cosden from Princeton, Catherine Schuman from ORNL, Mehrzad Samadi from NVIDIA, Stephen Lecler Olivier from Sandia National Labs, Douglas White from Oceanic, Fernanda Foertter from Next Silicon

Fall 2021 Highlights

Thu, 11 Mar 2021 00:00:00 +0000

Our lab GREW and we had to move out of the Computer Science Building. We are now at 111 S. Main Street, Newark, DE, 19716
Aaron Jarmusch and Nolan Baker present their posters at the SC21 Undergraduate Research Poster program
Eric Wright, our PhD student’s paper on “Refactoring the MPS/University of Chicago Radiative MHD (MURaM) model for GPU/CPU performance portability using OpenACC directives” Check out the full publication

New Grants and Awards

Thu, 11 Feb 2021 00:00:00 +0000

Eric Wright, Mauricio Ferrato, Sanhu Li, Vineeth Gutta, Matthew Leinhauser won the CIS Distinguished Graduate Student Award
Sunita Chandrasekaran chosen by the DOE to lead the Exascale Computing Project SOLLVE
Sunita Chandrasekaran has a new appointment - Computational Scientist at the Brookhaven National Laboratory

A Portable, High-Level Graph Analytics Framework Targeting Distributed, Heterogeneous Systems

Tue, 26 May 2020 11:04:20 +0200

Accelerating PPM_One

Tue, 26 May 2020 11:04:20 +0200

Exploring translation of OpenMP to OpenACC 2.5: Lessons learned

Tue, 26 May 2020 11:04:20 +0200

High-level Graph Analtics using MapReduce

Tue, 26 May 2020 11:04:20 +0200

NCAR's MURaM: Porting Radiative MHD Models to GPUs

Tue, 26 May 2020 11:04:20 +0200

NCAR’s MURaM: Porting the MPS/University of Chicago Radiative MHD Models to GPUs Using OpenACC

PI: Sunita Chandrasekaran
Student: Eric Wright
Collaborators from NCAR: Thomas Hauser, Cena Miller, Supreeth Suresh, Matthias Rempel
Collaborators from Max Planck Institute of Solar Society: Damien Przybylski

Funding Agency: National Center for Atmospheric Research (NCAR)/UCAR

Duration: 06/01/2018 – 09/30/2023

Project Summary:

NSF Eager

Tue, 26 May 2020 11:04:20 +0200

NSF EAGER: Measuring Real World Application Performance on Next-Generation Computing Systems

PI: Robert Henschel

Co-PIs: Sunita Chandrasekaran, Rudolf Eigemann

Student: Mayara Gimenes

Funding Agency: National Science Foundation (NSF)

Duration: 10/01/2018 – 03/31/2021

Project Summary:

This project aims to build an HPC2020 application-based benchmarking suite while working in collaboration with the Standard Performance Evaluation Corporation (SPEC) High Performance Group (HPG) harness.

NSF: PAW: Acceleration of Next Generation Sequence (NGS) Alignment

Tue, 26 May 2020 11:04:20 +0200

Acceleration of Next Generation Sequence (NGS) Alignment

PI: Sunita Chandrasekaran

Students: Sanhu Li

Collaborators: Nemours/Alfred I. duPont Hospital for Children: Erin Crowgey, Karl Franke

Duration: 06/30/2016 - work in progress

Project Summary:

The goal of this project is to design a scalable and portable sequence alignment tool that is both performance and memory efficient while not compromising on the accuracy, specificity or the sensitivity of the alignment. Such a tool is direly needed as it is evident from the COVID19 pandemic, that we are not ready to tackle the breakout of a novel virus. It has been 7 months since the breakout (at the time of writing this research statement, May 2020) and yet there is no vaccination. The purpose of this tool is to create a faster alignment tool than the current state-of-the-art gold-standard BWA-MEM that takes about 72 hours to finish a Whole Genome Sequence (WGS) alignment. My project has been scanning through several past and near-past literature to understand the complexities of creating a fast alignment tool. To that end for our tool, we have developed 3 novel algorithms: seeding, back-tracking and aligning. We are currently in the phase of integrating all steps and running several tests for accuracy, specificity and sensitivity. For benchmarking purposes, we use the NA12878 50X WGS sample from Illumina’s Genome in a Bottle along with the HG38 reference genomes.

NSF: SHF: PAW: Novel Functionality in Programming Models to Productively Abstract Wavefront Parallel Pattern

Tue, 26 May 2020 11:04:20 +0200

PI: Sunita Chandrasekaran

Students: Fabian Mora

Funding Agency: National Science Foundation (NSF)

Duration: 10/01/2018 – 09/30/2023

Project Summary:

The aim of this research project is to design novel high-level programming abstractions for complex parallel patterns in scientific applications as these patterns often require for the programmer to restructure the code thus spending hours to create a new codebase which can be both time-consuming and error-prone. To achieve this, we address the performance and portability questions at the algorithmic-level, programming framework-level and at the software design level. The studies are also suggestive of shortcomings in current programming models paving the way to developing novel insights towards high-level software abstractions for multi-use in different/diverse projects simultaneously.

OpenACC Acceleration of Molecular Dyanmics Chemical Shift Prediction

Tue, 26 May 2020 11:04:20 +0200

OpenACC NAS parallel benchmarks

Tue, 26 May 2020 11:04:20 +0200

OpenACC-Minisweep

Tue, 26 May 2020 11:04:20 +0200

Markdown Syntax Guide

Mon, 11 Mar 2019 00:00:00 +0000

This article offers a sample of basic Markdown syntax that can be used in Hugo content files, also it shows whether basic HTML elements are decorated with CSS in a Hugo theme.

Placeholder Text

Sat, 09 Mar 2019 00:00:00 +0000

Lorem est tota propiore conpellat pectoribus de pectora summo.

Math Typesetting

Fri, 08 Mar 2019 00:00:00 +0000

Mathematical notation in a Hugo project can be enabled by using third party JavaScript libraries.

Emoji Support

Tue, 05 Mar 2019 00:00:00 +0000

Emoji can be enabled in a Hugo project in a number of ways.

About

Thu, 28 Feb 2019 00:00:00 +0000

Written in Go, Hugo is an open source static site generator available under the Apache Licence 2.0. Hugo supports TOML, YAML and JSON data file types, Markdown and HTML content files and uses shortcodes to add rich content. Other notable features are taxonomies, multilingual mode, image processing, custom output formats, HTML/CSS/JS minification and support for Sass SCSS workflows.

Hugo makes use of a variety of open source projects including:

Hugo is ideal for blogs, corporate websites, creative portfolios, online magazines, single page applications or even a website with thousands of pages.

Sun, 06 Jan 2019 00:00:00 +0000

I spent 11 years of my life at University of Delaware obtaining my Bachelor’s Degree, Master’s Degree, and Doctorate. The best of those years were the nearly 3 years I had the pleasure of working with Prof. Sunita Chandrasekaran and the other students in her lab. She is smart, driven, and most importantly, she cares about her students. She pushes all of them to grow and learn, and she also cares about them on a personal level. She goes above and beyond to make sure that her students end up in places they want to be in their careers after leaving CRPL by leveraging all the connections she has made over the years in the field of HPC (which are extensive). In my case, I ended up working for my first-choice: NVIDIA.

Tue, 01 Jan 2019 00:00:00 +0000

I got my PhD and Masters from the University of Delaware. During this time I had the opportunity to work at the CRPL research group for several years. I obtained my master’s degree guided by Sunita and supported by my peers at CRPL. It was through this time that I really understood that a good graduate education experience can only be achieved through excellent academic advisors that have the ability to form strongly bonded research groups. Being a good professor and being able to create a strong research group goes beyond scientific relevance, and it is accompanied by a set of human and interpersonal skills that are often overlooked. Sunita’s research group shines for its academic excellence, but above all, for its human talent.

Tue, 01 Jan 2019 00:00:00 +0000

I loved my time in CRPL as an undergrad, it prepared me to succeed in a role where all my immediate teammates and developers I assist have PhDs. There is no distinction between who are the undergrads, masters, or PhDs when the group has gathered, all students excited by the research have a place at CRPL!

Contact Us

Mon, 01 Jan 0001 00:00:00 +0000

Location:

Our lab is located at FinTech Innovation Hub on Star Campus, University of Delaware.

Contact Form:

If you have any questions, or would like to reach out to us about something, please fill out the form below. We look forward to hearing from you!

Publications, Posters and Talks

Mon, 01 Jan 0001 00:00:00 +0000

For the full, year-by-year list, see the 📘Chronological List of Publications at the bottom.

📚 (Selected) Conferences and Workshops:

Jarmusch, Aaron, and Sunita Chandrasekaran. “Microbenchmarking NVIDIA’s Blackwell Architecture: An in-depth Architectural Analysis.” In proceedings at IPDPS 2026. arXiv preprint: 🔗 arxiv.org

Zachariah Sollenberger, Rahul Patel, Saieda Ali Zada, Sunita Chandrasekaran, LLM4VV: Evaluating Cutting-Edge LLMs for Generation and Evaluation of Directive-Based Parallel Programming Model Compiler Tests, to APPEAR in IEEE HiPC, Dec 2025 🔗 arxiv.org