# Per4ML Research Group — Data-Efficient AI/ML for HPC

> Dr. Tanzima Z. Islam is an Associate Professor of Computer Science at Texas State University. She leads the Per4ML laboratory, which develops data-efficient AI and ML methods for intelligent HPC systems. She holds the NSF CAREER Award (2025), the DOE Early Career Award (2022), and the R&D 100 Award (2019). She has secured over $4.6M in research funding from NSF, DOE, AMD, and national laboratories including LLNL, BNL, ORNL, and LBNL. She is also the co-founder of Bangladeshi Women in Computer Science and Engineering (BWCSE).

- Principal Investigator: Tanzima Z. Islam, Associate Professor, Computer Science, Texas State University
- Website: https://per4ml.github.io
- PI website: https://www.tanzimaislam.com
- Google Scholar: https://scholar.google.com/citations?user=q56uneAAAAAJ&hl=en
- Research funding secured: $4.6M+ (NSF, DOE, AMD, LLNL, BNL, ORNL, LBNL)

## Research Areas
The group works on, and welcomes collaborators and students in, the following topics:

- Multimodal Representation Learning
- Data-Efficient AI
- Foundation Models
- Heterogeneous Data
- Graphs
- Trees
- Tables
- Text
- Images
- Count Data
- Prediction
- Generation
- Counterfactual Decision Making
- Transfer Learning
- Agentic Systems
- Performance Optimization
- HPC
- Healthcare

## Current Team
- Tanzima Z. Islam (PI)
- Ankur Lahiri (PhD)
- Ashna Nawar Ahmed (PhD)
- Banooqa Banday (PhD)
- Chase Phelps (PhD)
- Elvis Fefey (PhD)
- Mohammad Zaeed (PhD)
- Nawshin Tabassum Tanny (PhD)
- Yugesh Bhattarai (PhD)
- Jaya Sravani (MS)

## Selected Projects
- CESAR Co-Design Project: Comparative Performance analysis — Performance
- CESAR Co-Design Project: Proxy Application Development and Validation — Performance
- Cache Bottleneck Characterization — Cache Analysis
- Data-driven decision making — Optimization
- ECRP+Fractal: Few-Shot Transfer Learning for Performance — Few-Shot Learning, HPC, Performance, Prediction, Transfer Learning
- ECRP: PerfGen: Synthesizing Performance using GenAI — GAN, GenAI, HPC, LLM, Machine Learning, Optimization, Performance
- ECRP: Performance-in-a-Graph (PinG) — HPC, Machine Learning, Performance
- MPI and MPI_T — Performance
- Performance Characterization — HPC, Optimization, Performance
- Performance-Aware Application Development — Machine Learning, Performance
- Ph.D. Thesis: Scalable Checkpoint/Restart — Fault Tolerance, Performance
- Power-Aware Resilience — HPC, Performance
- RECUP: Perf. Reproducibility — HPC, Performance, Reproducibility
- Reliable and Efficient Checkpointing for Cycle-sharing Systems — Fault Tolerance, Transfer Learning

## Publications
- Optimas: An Intelligent Analytics-Informed Generative AI Framework for Performance Optimization (2026). KDD. Authors: Zaeed, Mohammad, Islam, Tanzima Z, Indic, Vladimir
- COMPASS: A Unified Decision-Intelligence System for Navigating Performance Trade-off in HPC (2026). arXiv preprint arXiv:2604.22688. Authors: Lahiry, Ankur, Banday, Banooqa, Bhattarai, Yugesh, Zaeed, Mohammad, Islam, Tanzima Z
- Data-driven analysis to understand GPU hardware resource usage of optimizations (2026). The International Journal of High Performance Computing Applications. Authors: Islam, Tanzima Z, Schutte, Holland, Zaeed, Mohammad, Marathe, Aniruddha
- Attention-Informed Surrogates for Navigating Power-Performance Trade-offs in HPC (2025). NeurIPS ML for Systems Workshop. Authors: Nawar Ahmed, Ashna, Banday, Banooqa, Jones, Terry, Islam, Tanzima Z
- Report of the 2025 Workshop on Next-Generation Ecosystems for Scientific Computing: Harnessing Community, Software, and AI for Cross-Disciplinary Team Science (2025). arXiv preprint arXiv:2510.03413. Authors: McInnes, Lois Curfman, Arnold, Dorian, Balaprakash, Prasanna, Bernhardt, Mike, Cerny, Beth, Dubey, Anshu, Giles, Roscoe, Hood, Denice Ward, Leung, Mary Ann, Lopez-Marrero, Vanessa, others
- Reproducibility for HPC and Distributed Environments: Committees, Nondeterminism, Performance and Workflows (2025). Proceedings of the 3rd ACM Conference on Reproducibility and Replicability. Authors: Sirvent, Raül, Carratalá-Saez, Rocı́o, Gueroudji, Amal, Islam, Tanzima, Pouchard, Line, Taufer, Michela
- A Distributed Framework for Causal Modeling of Performance Variability in GPU Traces (2025). arXiv preprint arXiv:2510.18300. Authors: Lahiry, Ankur, Pokharel, Ayush, Banday, Banooqa, Ockerman, Seth, Gueroudji, Amal, Zaeed, Mohammad, Islam, Tanzima Z, Pouchard, Line
- Scalable GPU Performance Variability Analysis framework (2025). arXiv preprint arXiv:2506.20674. Authors: Lahiry, Ankur, Pokharel, Ayush, Ockerman, Seth, Gueroudji, Amal, Pouchard, Line, Islam, Tanzima Z
- HPC digital twins for evaluating scheduling policies, incentive structures and their impact on power and cooling (2025). Proceedings of the SC'25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis. Authors: Maiterth, Matthias, Brewer, Wesley H, Kuruvella, Jaya S, Dey, Arunavo, Islam, Tanzima Z, Kabir, Rashadul, Menear, Kevin, Duplyakin, Dmitry, Patki, Tapasya, Jones, Terry, others
- ModelX : A Novel Transfer Learning Approach Across Heterogeneous Datasets (2025). Proceedings of the 34th International Symposium on High-Performance Parallel and Distributed Computing. Authors: Dey, Arunavo, Antony, Neil, Dhakal, Aakash Raj, Thopalli, Kowshik, Thiagarajan, Jayaraman J., Patki, Tapasya, Marathe, Aniruddha, Scogland, Tom, Yeom, Jae-Seung, Islam, Tanzima. https://doi.org/10.1145/3731545.3731593
- Optimizing Deep Learning Inference on Heterogeneous Devices on the Edge: An ILP-Based Rate Monotonic Scheduling Approach (2025). 2025 IEEE Cloud Summit. Authors: Fefey, Elvis G, Islam, Tanzima. https://doi.org/10.1109/Cloud-Summit64795.2025.00011
- Structure-Aware Representation Learning for Effective Performance Prediction (2025). Concurrency and Computation: Practice and Experience. Authors: Ramadan, Tarek, Pinnow, Nathan, Phelps, Chase, Thiagarajan, Jayaraman J, Islam, Tanzima Z. https://doi.org/10.1002/cpe.70046
- xAMM: “Attention” to Details Improves Cross-Platform Prediction Accuracy (2025). 2025 IEEE 25th International Symposium on Cluster, Cloud and Internet Computing (CCGrid). Authors: Dhakal, Aakash Raj, Islam, Tanzima Z, Dey, Arunavo, Nichols, Daniel, Bhatele, Abhinav, Patki, Tapasya, Scogland, Tom, Yeom, Jae-Seung. https://doi.org/10.1109/CCGRID64434.2025.00067
- On the role of prompt construction in enhancing efficacy and efficiency of llm-based tabular data generation (2025). ICASSP 2025-2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Authors: Banday, Banooqa, Thopalli, Kowshik, Islam, Tanzima Z, Thiagarajan, Jayaraman J. https://doi.org/10.1109/ICASSP49660.2025.10888077
- Performance analysis and data reduction for exascale scientific workflows (2025). The International Journal of High Performance Computing Applications. Authors: Kelly, Christopher, Xu, Wei, Pouchard, Line C, Van Dam, Hubertus, Islam, Tanzima Z, Yoo, Shinjae, Kleese Van Dam, Kerstin
- WANDER: An Explainable Decision-Support Framework for HPC (2025). arXiv preprint arXiv:2506.04049. Authors: Lahiry, Ankur, Banday, Banooqa, Bhattarai, Yugesh, Islam, Tanzima Z
- Opal: A Modular Framework for Optimizing Performance using Analytics and LLMs (2025). arXiv preprint arXiv:2510.00932. Authors: Zaeed, Mohammad, Islam, Tanzima Z, Inđić, Vladimir
- PERFGEN: A Synthesis and Evaluation Framework for Performance Data using Generative AI (2024). 2024 IEEE 48th Annual Computers, Software, and Applications Conference (COMPSAC). Authors: Banday, Banooqa H, Islam, Tanzima Z, Marathe, Aniruddha. https://doi.org/10.1109/COMPSAC61105.2024.00035
- Relative performance prediction using few-shot learning (2024). 2024 IEEE 48th Annual Computers, Software, and Applications Conference (COMPSAC). Authors: Dey, Arunavo, Dhakal, Aakash, Islam, Tanzima Z, Yeom, Jae-Seung, Patki, Tapasya, Nichols, Daniel, Movsesyan, Alexander, Bhatele, Abhinav. https://doi.org/10.1109/COMPSAC61105.2024.00278
- Toward Efficient Deep Learning Inference: On-Node Heterogeneous Scheduling in Edge-Cloud Infrastructure (2024). 2024 IEEE Cloud Summit. Authors: Fefey, Elvis G, Islam, Tanzima. https://doi.org/10.1109/Cloud-Summit61220.2024.00019
- Reimagine Application Performance as a Graph: Novel Graph-Based Method for Performance Anomaly Classification in High-Performance Computing (2024). 2024 IEEE 48th Annual Computers, Software, and Applications Conference (COMPSAC). Authors: Phelps, Chase, Lahiry, Ankur, Islam, Tanzima Z, Pouchard, Line C. https://doi.org/10.1109/COMPSAC61105.2024.00041
- Characterize and compare the performance of deep learning optimizers in recurrent neural network architectures (2024). 2024 IEEE 48th Annual Computers, Software, and Applications Conference (COMPSAC). Authors: Zaeed, Mohammad, Islam, Tanzima Z, Inđić, Vladimir. https://doi.org/10.1109/COMPSAC61105.2024.00016
- Performance Characterization and Provenance of Distributed Task-based Workflows on HPC Platforms (2024). SC24-W: Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis. Authors: Gueroudji, Amal, Phelps, Chase, Islam, Tanzima Z, Carns, Philip, Snyder, Shane, Dorier, Matthieu, Ross, Robert B, Pouchard, Line C. https://doi.org/10.1109/SCW63240.2024.00254
- Automatic Parallelization of Cellular Automata for Heterogeneous Platforms (2023). 2023 IEEE 47th Annual Computers, Software, and Applications Conference (COMPSAC). Authors: Phelps, Chase, Islam, Tanzima. https://doi.org/10.1109/COMPSAC57700.2023.00055
- Signal processing based method for real-time anomaly detection in high-performance computing (2023). 2023 IEEE 47th Annual Computers, Software, and Applications Conference (COMPSAC). Authors: Dey, Arunavo, Islam, Tanzima, Phelps, Chase, Kelly, Christopher. https://doi.org/10.1109/COMPSAC57700.2023.00037
- Building the i (interoperability) of fair for performance reproducibility of large-scale composable workflows in recup (2023). 2023 IEEE 19th International Conference on e-Science (e-Science). Authors: Nicolae, Bogdan, Islam, Tanzima Z, Ross, Robert, Van Dam, Huub, Assogba, Kevin, Shpilker, Polina, Titov, Mikhail, Turilli, Matteo, Wang, Tianle, Kilic, Ozgur O, others
- Novel representation learning technique using graphs for performance analytics (2023). 2023 International Conference on Machine Learning and Applications (ICMLA). Authors: Ramadan, Tarek, Lahiry, Ankur, Islam, Tanzima Z
- Relative Performance Prediction using Semi-Supervised Learning (2023). Authors: Dey, A, Islam, T, Yeom, J, Patki, T, Nichols, D, Movsesyan, A, Bhatele, A
- Graph Based Anomaly Detection in Chimbuko: Feasible or Fallible? (2023). 2023 International Conference for High Performance Computing, Networking, Storage, and Analysis (SC'23). Authors: Phelps, Chase, Lahiry, Ankur, Islam, Tanzima Z, Kelly, Christopher
- Interactive Visual Analysis Tool for Anomaly Provenance Data (2022). Authors: Guite, A, Islam, T, Kelley, C, Xu, W. https://sc22.supercomputing.org/proceedings/tech_poster/poster_files/rpost159s3-file3.pdf
- Analysis and Visualization of Important Performance Counters To Enhance Interpretability of Autotuner Output (2022). Authors: Zaeed, Mohammad, Islam, Tanzima Z, Cho, Younghyun, Li, Xiaoye Sherry, Luo, Hengrui, Liu, Yang. https://sc22.supercomputing.org/proceedings/tech_poster/poster_files/rpost183s3-file3.pdf
- LIBNVCD: An Extendable and User-friendly Multi-GPU Performance Measurement Tool (2022). 2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC). Authors: Schutte, Holland, Phelps, Chase, Marathe, Aniruddha, Islam, Tanzima. https://doi.org/10.1109/COMPSAC54236.2022.00019
- Novel proposals for FAIR, automated, recommendable, and robust workflows (2022). 2022 IEEE/ACM Workshop on Workflows in Support of Large-Scale Science (WORKS). Authors: Abhinit, Ishan, Adams, Emily K, Alam, Khairul, Chase, Brian, Deelman, Ewa, Gorenstein, Lev, Hudson, Stephen, Islam, Tanzima, Larson, Jeffrey, Lentner, Geoffrey, others
- Challenges for implementing fair digital objects with high performance workflows (2022). Research Ideas and Outcomes. Authors: Pouchard, Line, Islam, Tanzima, Nicolae, Bogdan
- Low Rank Approximation Methods for Identifying Impactful Pairwise Protein Mutations (2022). Algorithms and Methods in Structural Bioinformatics. Authors: Daw, Chris, Cruz, Brian Barragan, Majeske, Nicholas, Jagodzinski, Filip, Islam, Tanzima, Hutchinson, Brian. https://doi.org/10.1007/978-3-031-05914-8_4
- Comparative code structure analysis using deep learning for performance prediction (2021). 2021 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS). Authors: Ramadan, Tarek, Islam, Tanzima Z, Phelps, Chase, Pinnow, Nathan, Thiagarajan, Jayaraman J. https://doi.org/10.1109/ISPASS51385.2021.00032
- HPC@ SCALE: a hands-on approach for training next-gen HPC software architects (2021). 2021 IEEE 28th International Conference on High Performance Computing, Data and Analytics Workshop (HiPCW). Authors: Islam, Tanzima Z, Phelps, Chase. https://doi.org/10.1109/HiPCW54834.2021.00011
- College life is hard!-shedding light on stress prediction for autistic college students using data-driven analysis (2021). 2021 IEEE 45th Annual Computers, Software, and Applications Conference (COMPSAC). Authors: Islam, Tanzima Z, Liang, Philip Wu, Sweeney, Forest, Pragner, Cody, Thiagarajan, Jayaraman J, Sharmin, Moushumi, Ahmed, Shameem
- FILCIO: Application agnostic I/O aggregation to scale scientific workflows (2021). 2021 IEEE 45th Annual Computers, Software, and Applications Conference (COMPSAC). Authors: Jensen, Quentin, Jagodzinski, Filip, Islam, Tanzima
- Towards aggregation based i/o optimization for scaling bioinformatics applications (2020). 2020 IEEE 44th Annual Computers, Software, and Applications Conference (COMPSAC). Authors: Stratton, Jack, Albert, Michael, Jensen, Quentin, Ismailov, Max, Jagodzinski, Filip, Islam, Tanzima
- Performance optimality or reproducibility: that is the question (2019). Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis. Authors: Patki, Tapasya, Thiagarajan, Jayaraman J, Ayala, Alexis, Islam, Tanzima Z
- Toward a programmable analysis and visualization framework for interactive performance analytics (2019). 2019 IEEE/ACM International Workshop on Programming and Performance Visualization Tools (ProTools). Authors: Islam, Tanzima, Ayala, Alexis, Jensen, Quentin, Ibrahim, Khaled
- Paddle: Performance analysis using a data-driven learning environment (2018). 2018 IEEE International Parallel and Distributed Processing Symposium (IPDPS). Authors: Thiagarajan, Jayaraman J, Anirudh, Rushil, Kailkhura, Bhavya, Jain, Nikhil, Islam, Tanzima, Bhatele, Abhinav, Yeom, Jae-Seung, Gamblin, Todd
- Low Rank Smoothed Sampling Methods for Identifying Impactful Pairwise Mutations (2018). Proceedings of the 2018 ACM International Conference on Bioinformatics, Computational Biology, and Health Informatics. Authors: Majeske, Nicholas, Jagodzinski, Filip, Hutchinson, Brian, Islam, Tanzima
- Automatic Generation of Mixed-Precision Programs (2018). Authors: Moody, Logan, Pinnow, Nathan, Lam, Michael O, Menon, Harshitha, Schordan, Markus, Lloyd, G Scott, Islam, Tanzima
- Metakv: A key-value store for metadata management of distributed burst buffers (2017). 2017 IEEE International Parallel and Distributed Processing Symposium (IPDPS). Authors: Wang, Teng, Moody, Adam, Zhu, Yue, Mohror, Kathryn, Sato, Kento, Islam, Tanzima, Yu, Weikuan
- A new proxy application for compressible multiphase turbulent flows (2017). Sustainable Computing: Informatics and Systems. Authors: Banerjee, Tania, Hackl, Jason, Shringarpure, Mrugesh, Islam, Tanzima, Balachandar, S, Jackson, Thomas, Ranka, Sanjay
- A machine learning framework for performance coverage analysis of proxy applications (2016). SC'16: Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis. Authors: Islam, Tanzima Z, Thiagarajan, Jayaraman J, Bhatele, Abhinav, Schulz, Martin, Gamblin, Todd
- Exploring the MPI tool information interface: features and capabilities (2016). The International Journal of High Performance Computing Applications. Authors: Islam, Tanzima, Mohror, Kathryn, Schulz, Martin
- I/O aware power shifting (2016). 2016 IEEE International Parallel and Distributed Processing Symposium (IPDPS). Authors: Lee, Savoie, Lowenthal, David K, De Supinski, Bronis R, Islam, Tanzima, Mohror, Kathryn, Rountree, Barry, Schulz, Martin
- Cmt-bone—a proxy application for compressible multiphase turbulent flows (2016). 2016 IEEE 23rd International Conference on High Performance Computing (HiPC). Authors: Banerjee, Tania, Hackl, Jason, Shringarpure, Mrugesh, Islam, Tanzima, Balachandar, S, Jackson, Thomas, Ranka, Sanjay
- Fault Tolerance Assistant (FTA): An Exception Handling Programming Model for MPI Applications (2016). Authors: Fang, A, Laguna, I, Sato, K, Islam, T, Mohror, K
- Fault Tolerance Assistant (FTA): An Exception Handling Approach for MPI Programs (2015). ExaMPI15 Exascale MPI at Supercomputing 2015 (SC15). Authors: Fang, Aiman, Laguna, Ignacio, Sato, Kento, Islam, Tanzima, Mohror, Kathryn
- Reliable and efficient distributed checkpointing system for grid environments (2014). Journal of grid computing. Authors: Islam, Tanzima Zerin, Bagchi, Saurabh, Eigenmann, Rudolf
- Exploring the Capabilities of the New MPI_T Interface (2014). Proceedings of the 21st European MPI Users' Group Meeting. Authors: Islam, Tanzima, Mohror, Kathryn, Schulz, Martin
- Lossy Compression for Checkpointing: Fallible or Feasible? (2014). International Conference for High Performance Computing, Networking, Storage, and Analysis. Authors: Ni, Xiang, Kale, Laxmikant, Islam, Tanzima, Mohror, Kathryn, Moody, Adam
- Batchsubmit: a high-volume batch submission system for earthquake engineering simulation (2014). Concurrency and Computation: Practice and Experience. Authors: Mohan, Anup, Hacker, Thomas, Rodgers, Gregory P, Islam, Tanzima
- XSBench -- the development and verification of a performance abstraction for Monte Carlo reactor analysis (2014). Proceedings of PHYSOR 2014, Kyoto, Japan, Sep 28 -- Oct 3, 2014. Authors: Tramm, John R, Siegel, Andrew R, Islam, Tanzima, Schulz, Martin
- Reliable and Scalable Checkpointing Systems (2013). Authors: Islam, Tanzima Zerin
- McrEngine: a scalable checkpointing system using data-aware aggregation and compression (2012). Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis. Authors: Islam, Tanzima Zerin, Mohror, Kathryn, Bagchi, Saurabh, Moody, Adam, de Supinski, Bronis R., Eigenmann, Rudolf
- Eunomia: A fast and collaborative interference avoidance technique for wireless medical devices (2011). Authors: Hossain, Mohammad S, Islam, Tanzima Zerin, Wei, Jinliang, Raghunathan, Vijay, Bagchi, Saurabh
- FALCON: a system for reliable checkpoint recovery in shared grid environments (2009). Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis. Authors: Islam, Tanzima Zerin, Bagchi, Saurabh, Eigenmann, Rudolf. https://doi.org/10.1145/1654059.1654110
- Gpnocsim - A General Purpose Simulator for Network-On-Chip (2007). 2007 International Conference on Information and Communication Technology. Authors: Hossain, Hemayet, Ahmed, Mostak, Al-Nayeem, Abdullah, Islam, Tanzima Zerin, Akbar, Md. Mostofa
- Development of an Improved General Purpose Simulator for Network-On-Chip (2006). Authors: Al-Nayeem, Abdullah, Islam, Tanzima Zerin

## Alumni / Past Members
- Tarek Ramadan (M.Sc., Oracle)
- Russell Hernandez Ruiz (B.Sc., TxState)
- Ethan Greene (B.Sc., TxState)
- Holland Schutte (B.Sc., WWU)
- Gian-Carlo DeFazio (B.Sc., currently at LLNL)
- Nathan Pinnow (M.Sc., currently at LLNL)
- Nicholas Majeske (M.Sc., currently Ph.D. student at Indiana University)
- Anna Zivkovik (B.Sc., WWU)
- Jack Stratton (B.Sc., WWU)
- David Smith (B.Sc., WWU)
- Tony Dinh (B.Sc., WWU)
- Trevor Marcus (B.Sc., WWU)
- Chloe Dawson (B.Sc., WWU)
- Alexis Ayala (B.Sc., graduate student at WWU)
- Quentin Jensen (B.Sc., graduate student at WWU)
- Philip Wu Liang (B.Sc., WWU)
- Forest Sweeney (B.Sc., WWU)
- Cody Pragner (B.Sc., WWU)

## Collaborators
- Khaled Ibrahim (LBNL)
- Yang Liu (LBNL)
- Jae-Seung Yeom (LLNL)
- Jayaraman J. Thiagarajan (Former: LLNL, Current: Apple)
- Barry Rountree (LLNL)
- Aniruddha Marathe (LLNL)
- Kathryn Mohror (LLNL)
- Tapasya Patki (LLNL)
- Kerstin Kleese Van Dam (BNL)
- Line Pouchard (Sandia NL)
- Bogdan Nicolae (ANL)
- Rob Ross (ANL)
- Terry Jones (ORNL)

## For prospective students & collaborators
The lab welcomes PhD applicants, MS students, and undergraduate researchers interested in the research areas above. Reach out via the PI's website: https://www.tanzimaislam.com