, Ph.D. E-mail: kokrommi@gmail.com Phone: (+1) 540-394-0522 Web: www.cs.vt.edu/~kokrommy w LinkedIn: www.linkedin.com/in/krommydas EDUCATION Virginia Polytechnic Institute and State University (Virginia Tech), Department of Computer Science (August 2010 May 2017): Ph.D. Degree Thesis: Towards Enhancing Performance, Programmability, and Portability in Heterogeneous Computing GPA: 3.8/4 Advisor: Dr. Wu-chun Feng Virginia Polytechnic Institute and State University (Virginia Tech), Department of Computer Science (August 2010 December 2013): Master s Degree (non-thesis) GPA: 3.8/4 Advisor: Dr. Wu-chun Feng University of Thessaly (Greece), Department of Computer and Communications Engineering (September 2004 - June 2010): B.S. Diploma (5-year program, Thesis) GPA: 8.24/10 (Top 3% of graduating class) Thesis: Implementation of the AVS Video Decoder in Reconfigurable Logic Co-advisors: Dr. Nikolaos Bellas, Dr. Christos D. Antonopoulos RESEARCH INTERESTS High performance computing, Systems, Parallel computer architectures, Programming languages, Bioinformatics, Image and video processing TECHNICAL STRENGTHS Programming: C, C++, FORTRAN, Java, Perl, Shell scripting, SQL, HTML, PHP, JavaScript, SWIFT, MIPS assembly Parallel Programming: OpenMP, MPI, CUDA, OpenCL, Pthreads, MMX/SSE/AVX SIMD extensions Profiling & Debugging Tools: Intel VTune, NVIDIA Visual Profiler, AMD CodeXL, gprof, Intel Advisor, gdb Other Software & Tools: MS Visual Studio, Latex, SVN, GIT, Flex/Bison, MS Office 1
WORK EXPERIENCE Intel: Software Engineer, February 2017 now o Optimizing high-performance computing (HPC) parallel applications (quantum chemistry, computational cosmology) on Intel platforms (CPU, network fabric, I/O). Research on architectural innovations for future Intel processors based on HPC domain algorithm insights. Intel: Software Engineering Graduate Intern, August - December 2015 o Extended prior research prototype of GLAF, a visual-programming environment, for domain scientists to automate (parallel) OpenCL code generation. Enabled seamless portability of existing programs to GPU and FPGA platforms and enhanced programmability and functional/performance portability. Intel: Software Engineering Graduate Intern, May - December 2014 o Designed and implemented aspects of GLAF, an all-encompassing visual programming environment. Key elements are its intuitive programming interface for scientists, researchers and engineers, and its serial and parallel code generation (C, FORTRAN) and optimization back-ends. Virginia Tech: Graduate Teaching Assistant (CS 4504: Computer Architecture - Spring 2014, Spring 2016 and CS 4234: Parallel Computing - Fall 2016) o Assisted with teaching (multiple guest lectures), and material creation (lecture presentations, homework problems, programming assignments, semester-long projects, exams). Led review sessions, and did grading. Virginia Tech: Graduate Research Assistant (Synergy Lab), August 2010 December 2016 (except periods noted above) o Extended AVS video decoder optimizations and explored performance trade-offs in large-scale multiprocessors (experimental Intel Many-core Testing Lab). Addressed scalability issues and devised algorithmic remedies to overcome architectural bottlenecks and load imbalance. o Investigated the suitability and performance of fine-grained, hybrid parallelization approaches for protein sequence alignment (BLAST algorithm) on the GPU. Proposed and prototyped compressive genomics approaches for storage reduction and performance improvements. o Examined performance and portability aspects of the OpenCL Dwarfs benchmark suite on architecturally diverse, heterogeneous parallel platforms (CPU, APU, GPU, FPGA and Intel MIC). Revamped OpenCL Dwarfs by extending robustness, benchmarks coverage and benchmarking capabilities, and enabling uniformity of cross-platform optimization levels. o Mapped, optimized, and characterized molecular modeling (electrostatic surface potential calculation) algorithms across modern multi- and many-core architectures. Conducted a quantitative and qualitative study on the performance, programmability and portability trade-offs. o Proposed and implemented a novel motif-based benchmarking methodology for rapid evaluation of composite/fused heterogeneous architectures. Applied the proposed methodology to a diverse set of candidate platforms and provided projected performance insights. 2
o Enhanced robustness of our GLAF visual-programming framework to support automatic parallelization and optimization of large-scale FORTRAN legacy applications. Applied on NASA weather modeling and computational fluid dynamics algorithms (NASA-funded project). University of Thessaly: Undergraduate Researcher, Spring 2009 Spring 2010 o Algorithmic mapping and performance optimization of the AVS Video decoder on multicore CPUs. Exploited x86 SIMD capabilities and parallelized using macroblock-level dynamic multithreading schemes. Achieved real-time highdefinition video decoding. o Implemented and evaluated AVS video decoder kernels on a Xilinx ML405 FPGA with embedded PowerPC CPU. Created soft core IP for custom motion compensation peripheral and enabled integration within an embedded Xilinx PowerPC-based Linux distribution. PEER-REVIEWED PUBLICATIONS Krommydas, K.; Feng, W-C.; Telescoping Architectures: Evaluating Next Generation Heterogeneous Computing, 2016 IEEE International Conference on High Performance Computing, Data, and Analytics (HiPC), 19-22 December 2016 Krommydas, K.; Sasanka, R.; Feng, W-C.; Bridging the FPGA Performance- Programmability Gap via Automatic OpenCL Code Generation and Tuning, 2016 IEEE International Conference on Application-specific Systems, Architectures and Processors (ASAP), 6-8 July 2016 Krommydas, K.; Feng, W-C.; Antonopoulos, C.D.; Bellas, N.; OpenDwarfs: Characterization of Dwarf-based Benchmarks on Fixed and Reconfigurable Architectures, Journal of Signal Processing Systems, Springer, October 2015 Krommydas, K.; Sasanka, R.; Feng, W-C.; GLAF: A Visual Programming and Auto-tuning Framework for High-performance Computing, 2015 ACM International Conference on Supercomputing (ICPP), 1-4 September 2015 Kalidas, R.; Daga, M.; Krommydas, K.; Feng, W-C.; On the Performance, Energy, and Power of Data-Access Methods in Heterogeneous Computing Systems, 2015 IEEE International Parallel and Distributed Processing Symposium Workshops High-Performance, Power-Aware Computing (HPPAC), 25-29 May 2015 Feng, W-C.; Krommydas, K.; Zhang L.; CentroidBLAST: Accelerating Sequence Search via Clustering, 7 th International Conference on Bioinformatics and Computational Biology (BICOB), 9-11 March 2015 Krommydas, K.; Feng, W-C.; Owaida, M.; Antonopoulos, C.D.; Bellas, N.; On the Characterization of OpenCL Dwarfs on Fixed and Reconfigurable Platforms, 2014 IEEE International Conference on Application-specific Systems, Architectures and Processors (ASAP), 18-20 June 2014 (Best paper finalist Top 3) Krommydas, K.; Scogland, T.; Feng W-C.; On the Programmability and Performance of Heterogeneous Architectures, 2013 IEEE International Conference on Parallel and Distributed Systems (ICPADS), 15-18 December 2013 3
Krommydas, K.; Antonopoulos, C.D.; Bellas, N.; Feng, W-C.; AVS Video Decoder on Multicore Systems: Optimizations and Tradeoffs, 2011 IEEE International Conference on Multimedia and Expo (ICME), 11-15 July 2011 Krommydas, K.; Tsoublekas, G.; Antonopoulos, C.D.; Bellas, N.; Mapping and Optimization of the AVS Video Decoder on a High Performance Chip Multiprocessor, 2010 IEEE International Conference on Multimedia and Expo (ICME), 19-23 July 2010 INVITED TALKS/PRESENTATIONS GLAF: Parallel Programming for the Masses, NASA Langley Research Center, Hampton, VA, May 10 2017 Tutorial: Accessing and Using the GLAF Prototype, NASA Langley Research Center, Hampton, VA, September 30 2016 Performance, Programmability, and Portability in Heterogeneous Computing: A Multi-Faceted Perspective and Case Study, Computer Science Department Seminar - Old Dominion University (ODU), Norfolk, VA, March 4 2016 HONORS AND AWARDS ICTAS Doctoral Scholars Program Award, Virginia Tech (2010-2014) Gerondelis Foundation Scholarship, USA (2011) Technical Chamber of Greece Top Student Award (awarded yearly to the top graduating student of each Polytechnic department), Greece (2010) PROFESSIONAL SERVICE AND ACTIVITIES Referee of Publications: o ACM Transactions of Reconfigurable Technology and Systems (TRETS) o IEEE Transactions on Very Large Scale Integration Systems (TVLSI) o ACM Transactions on Multimedia Computing Communications and Applications (TOMCCAP) o Design Automation Conference (DAC) o IEEE Access Journal o International Journal of High Performance Computing Applications (IJHPCA) Sage o Journal on Concurrency and Computation: Practice and Experience - Wiley o IETE Technical Review Journal Taylor & Francis o International Workshop on Accelerators and Hybrid Exascale Systems (AsHES) IEEE IPDPS Workshop (IPDPSW) 4
o Morgan-Kaufmann Book Proposals Reviewer ( High-performance Parallelism Pearls II Multicore and Manycore Approaches, Intel Xeon Phi Processor (Knights Landing) High Performance Programming ) Technical Program Committee/Conference Organization Participation: o Vice-Chair of Poster Session, 47 th Processing (ICPP 2018) International Conference on Parallel o Technical Program Committee Member, 7 th Workshop on Parallel Programming Models (IPDPS 2018 Workshop) Professional Membership: o IEEE Administrative Service: o President for the Virginia Tech Hellenic Student Association (2013-2015) o Public Relations Officer for the Virginia Tech Hellenic Student Association (2011-2013) o Graduate Student Committee for the Virginia Tech College of Engineering (2010-2011) o Local Organization Committee for the 3rd National Pan-Hellenic Electrical & Computer Engineering Students Conference - University Of Thessaly branch (2009) o Student Representative at the General Board of the Computer and Communication Engineering Dept. (2007-2008) 5