Faculty Directory

Biography

Mehmet E. Belviranli received his Ph.D. degree in Computer Science from University of California, Riverside in 2016. Prior to joining Mines, he continued his research and mentoring activities at Oak Ridge National Laboratory as Staff Computer Scientist.  Mehmet's main research interests are targeted at increasing resource utilization in heterogeneous architectures. He developed runtime systems, scheduling algorithms, analytical models, extended memory spaces, programming abstractions and OS & architecture level support to make heterogeneous computing more efficient.

His work has been published in major HPC and architecture related conferences including MICRO, PPoPP, SC, ICS, PACT and DATE.

Personal Homepage

Education

• Ph.D. Computer Science, University of California, Riverside, 2016
• M.S Computer Science, Bilkent University, 2009
• B.S. Computer Science, Bilkent University, 2006

Research

• Diversely heterogeneous architectures
• Runtime Systems
• Analytical performance & resource modeling
• Parallel Programming Paradigms
• Autonomous Computing
• Systems Organization for Machine Learning
• Cyber-physical systems
• Deep learning acceleration
• Parallel programming paradigms
• Edge-cloud platforms

Publications

[ISPASS'26] Hibiscus: End-to-end Architectural Simulation Framework for Hybrid SFQ/CMOS-Memory Compute Systems (Tool or Benchmark Paper)
Marsala, Ryan; Yerzhan, Mustafa; Tangella, Prabhat; Sonji, Muhammed; Michelogiannakis, George; Kose, Selcuk; Jog, Adwait; Belviranli, Mehmet E
Proceedings of the 2026 IEEE International Symposium on Performance Analysis of Systems and Software

[ICS'25] HARNESS: Holistic Resource Management for Diversely Scaled Edge Cloud Systems
Dagli, Ismet; Davis, Justin; Belviranli, Mehmet E
Proceedings of the 2025 International Conference on Supercomputing

[DATE'25] MC3: Memory Contention-based Covert Channel Communication on Shared DRAM System-on-Chips
Dagli, Ismet; Crea, James; Seckiner, Soner; Xu, Yuanchao; Köse, Selçuk; Belviranli, Mehmet E
2025 Design, Automation & Test in Europe Conference & Exhibition

[ICS'24] Scheduling for Cyber-Physical Systems with Heterogeneous Processing Units under Real-World Constraints
McGowen, Justin; Dagli, Ismet; Dantam, Neil T.; Belviranli, Mehmet E
Proceedings of the 2024 International Conference on Supercomputing

[PPoPP'24] Shared Memory-contention-aware Concurrent DNN Execution for Diversely Heterogeneous SoCs
Dagli, Ismet;Belviranli, Mehmet E
Proceedings of ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming.

[DATE'24] Context-aware Multi-Model Object Detection for Diversely Heterogeneous Compute Systems
Davis, Justin; Belviranli, Mehmet E; (Outstanding Paper Award in Autonomous Systems Design)
2024 Design, Automation & Test in Europe Conference & Exhibition

[PACT'22] Optimizing Regular Expressions via Rewrite-Guided Synthesis
Jedidiah McClurg; Miles Claver; Jackson Garner; Jake Vossen; Jordan Schmerge; Belviranli, Mehmet E
31st International Conference on Parallel Architectures and Compilation Techniques

[DAC'22] AxoNN: Energy-Aware Execution of Neural Network Inference on Multi-Accelerator Heterogeneous SoCs
Ismet Dagli; Alexander Cieslewicz; Jedidiah McClurg; Belviranli, Mehmet E
59th ACM/IEEE Design Automation Conference

[MICRO'21] PCCS: Processor-Centric Contention-aware Slowdown Model for Heterogeneous SoCs
Yuanchao Xu; Belviranli, Mehmet E; Xipeng Shen; Vetter, Jeffrey S;
54th Annual IEEE/ACM International Symposium on Microarchitecture

[PARCO] A computational-graph partitioning method for training memory-constrained DNNs
Fareed Qararyah, Mohamed Wahib, Doğa Dikbayır, Mehmet Esat Belviranli , Didem Unat
Parallel Computing, 2021, 102792, ISSN 0167-8191

[PACT'20] MEPHESTO: Modeling Energy-Performance in Heterogeneous SoCs and Their Trade-Offs
Monil, Mohammad; Belviranli, Mehmet E; Lee, Seyong; Vetter, Jeffrey S; Malony, Allen;
2020 International Conference on Parallel Architectures and Compilation Techniques

[DATE'19] FLAME: Graph-based Hardware Representations for Rapid and Precise Performance Modeling
Belviranli, Mehmet E; Vetter, Jeffrey S;
2019 Design, Automation & Test in Europe Conference & Exhibition

[SC'18] DRAGON: Breaking GPU Memory Capacity Limits with Direct NVM Access
Markthub, Pak; Belviranli, Mehmet E; Lee, Seyong; Vetter, Jeffrey S; Matsuoka, Satoshi;
Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis

[HPEC'18] Designing Algorithms for the EMU Migrating-threads-based Architecture
Belviranli, Mehmet E; Lee, Seyong; Vetter, Jeffrey S; (Best Paper Finalist)
2018 IEEE High Performance Extreme Computing Conference

[PPoPP'18] Juggler: A Dependency-Aware Task Based Execution Framework for GPUs
Belviranli, Mehmet E; Lee, Seyong; Vetter, Jeffrey S; Bhuyan, Laxmi N;
Proceedings of ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming.

[MICRO'17] Wireframe: supporting data-dependent parallelism through dependency graph execution in GPUs.
Abdolrashidi, Amir A; Tripathy, Devashree; Belviranli, Mehmet E; Bhuyan, Laxmi N;
Proceedings of the 50th Annual IEEE/ACM International Symposium on Microarchitecture

[ICS'16] CuMAS: Data transfer aware multi-application scheduling for shared GPUs
Belviranli, Mehmet E; Khorasani, Farzad; Bhuyan, Laxmi N; Gupta, Rajiv;
Proceedings of the 2016 International Conference on Supercomputing

[ICS'15] Peerwave: Exploiting wavefront parallelism on GPUs with peer-sm synchronization
Belviranli, Mehmet E; Deng, Peng; Bhuyan, Laxmi N; Gupta, Rajiv; Zhu, Qi;
Proceedings of the 2015 International Conference on Supercomputing

[PACT'15] Stadium hashing: Scalable and flexible hashing on GPUs
Khorasani, Farzad; Belviranli, Mehmet E; Gupta, Rajiv; Bhuyan, Laxmi N;
International Conference on Parallel Architecture and Compilation

[TACO'13] A dynamic self-scheduling scheme for heterogeneous multiprocessor architectures
Belviranli, Mehmet E; Bhuyan, Laxmi N; Gupta, Rajiv;
ACM Transactions on Architecture and Code Optimization 2013