v(LastName) google.com

I am a compiler developer at Google in the Bay Area, developing a new auto-vectorizer in upstream LLVM based on Sandbox IR.

Before joining Google I was a compiler developer at Intel in Santa Clara, and before that I was workng on compilers at the University of Cambridge.

I did my Ph.D. in compilers at the School of Informatics, University of Edinburgh. My thesis was on code generation for scalable low-power VLIW processors in GCC.

Publications (DBLP) (Google Scholar)

Research Area	Publications
Auto-Vectorization	[CGO'19] Super-Node SLP: Optimized Vectorization for Code Sequences Containing Operators and Their Inverse Elements Vasileios Porpodas, Rodrigo C. O. Rocha, Evgueni Brevnov, Luís F. W. Góes, Timothy Mattson. Intl. Symp. on Code Generation and Optimization (CGO), Feb 2019. pdf bib slides LLVM Patch Poster@EuroLLVM'19 [LCPC'19] PostSLP: Cross-Region Vectorization of Fully or Partially Vectorized Code Vasileios Porpodas and Pushkar Ratnalikar. Intl. Wksp. on Languages and Compilers for Parallel Computing (LCPC), Oct 2019. pdf bib slides [PACT'18] VW-SLP: Auto-Vectorization with Adaptive Vector Width Vasileios Porpodas, Rodrigo C. O. Rocha and Luís F. W. Góes. Intl. Conference on Parallel Architectures and Compilation Techniques (PACT), Nov 2018. pdf bib slides demo video(LLVM-DEV'18) [CGO'18] Look-Ahead SLP: Auto-Vectorization in the Presence of Commutative Operations Vasileios Porpodas, Rodrigo C. O. Rocha and Luís F. W. Góes. Intl. Symp. on Code Generation and Optimization (CGO), Feb 2018. pdf bib slides LLVM SLP Patch VPlan Patch (third-party) Video@EuroLLVM'18 [PACT'17] SuperGraph SLP Auto-Vectorization. Vasileios Porpodas. Intl. Conference on Parallel Architectures and Compilation Techniques (PACT), Sep 2017. pdf bib slides [PACT'15] Throttling Automatic Vectorization: When Less Is More. Vasileios Porpodas, and Timothy M. Jones. Intl. Conference on Parallel Architectures and Compilation Techniques (PACT), Oct 2015. pdf bib slides LLVM patch (third-party) video(LLVM'15) [CGO'15] PSLP: Padded SLP Automatic Vectorization. Vasileios Porpodas, Alberto Magni and Timothy M. Jones. Intl. Symp. on Code Generation and Optimization (CGO), Feb 2015. pdf bib slides video(EuroLLVM'15) award@SCOPES'15
Scalar Optimizations	[CC'20] Vectorization-aware loop unrolling with seed forwarding Rodrigo C. O. Rocha, Vasileios Porpodas, Pavlos Petoumenos, Luís F. W. Góes, Zheng Wang, Murray Cole and Hugh Leather. Intl. Conference on Compiler Construction (CC), 2020. pdf bib
Parallel Data Structures	[ICS'16]Lynx: Using OS and Hardware Support for Fast Fine-Grained Inter-Core Communication. Konstantina Mitropoulou, Vasileios Porpodas, Xiaochun Zhang and Timothy M. Jones. International Conference on Supercomputing (ICS), June 2016. pdf bib slides source code in C and C++
Instruction Scheduling	[LCPC'13] Aligned Scheduling: Cache-efficient Instruction Scheduling for VLIW Processors. Vasileios Porpodas and Marcelo Cintra. Intl. Wksp. on Languages and Compilers for Parallel Computing (LCPC), Sep 2013. pdf bib slides [CASES'13] CAeSaR: unified Cluster-Assignment Scheduling and communication Reuse for clustered VLIW processors. Vasileios Porpodas and Marcelo Cintra. Intl. Conf. on Compilers Architecture and Synthesis for Embedded Systems (CASES), Sep 2013. pdf bib slides [LCTES'13] LUCAS: Latency-adaptive Unified Cluster Assignment and instruction Scheduling. Vasileios Porpodas and Marcelo Cintra. Conf. on Languages, Compilers and Tools for Embedded Systems (LCTES), Jun 2013. pdf bib slides [LCPC'12] UCIFF: Unified Cluster Assignment, Instruction Scheduling, and Fast Frequency Selection for Heterogeneous Clustered VLIW Cores. Vasileios Porpodas and Marcelo Cintra. Intl. Wksp. on Languages and Compilers for Parallel Computing (LCPC), Sep 2012. pdf bib slides
Fault Tolerance	[arXiv] ZOFI: Zero Overhead Fault Injection Tool for Fast Transient Fault Coverage Analysis Vasileios Porpodas. arXiv report, Jun 2019. pdf bib Download tool / Source code [CASES'16] COMET: Communication-Optimised Multi-threaded Error-detection Technique. Konstantina Mitropoulou, Vasileios Porpodas and Timothy M. Jones. Intl. Conference on Compilers, Architecture and Synthesis for Embedded Systems (CASES), Oct 2016. pdf bib [LCPC'13] DRIFT: Decoupled compileR-based Instruction-level Fault-Tolerance. Konstantina Mitropoulou, Vasileios Porpodas and Marcelo Cintra. Intl. Wksp. on Languages and Compilers for Parallel Computing (LCPC), Sep 2013. pdf bib slides [IPDPS'13] CASTED: Core-Adaptive Software Transient Error Detection for Tightly Coupled Cores. Konstantina Mitropoulou, Vasileios Porpodas and Marcelo Cintra. Intl. Parallel & Distributed Processing Symposium (IPDPS), May 2013. pdf bib
HW / Architecture	[HPCA'12] Cooperative Partitioning: Energy-Efficient Cache Partitioning for High-Performance CMPs. Karthik T. Sundararajan, Vasileios Porpodas, Timothy M. Jones, Nigel P. Topham and Björn Franke. Intl. Symp. on High-Performance Computer Architecture (HPCA), February 2012. pdf bib [JSPS] Decoupled Processors Architecture for Accelerating Data Intensive Applications using Scratch-Pad Memory Hierarchy. A. Milidonis, N. Alachiotis, V. Porpodas, H. Michail, G. Panagiotakopoulos, A.P. Kakarountas, and C.E. Goutis. Journal of Signal Processing Systems (JSPS), pages 1-16, 2009. bib [CDT] Low-power architecture with scratch-pad memory for accelerating embedded applications with run-time reuse. A. Milidonis, V. Porpodas, N. Alachiotis, A.P. Kakarountas, H. Michail, G. Panagiotakopoulos, and CE Goutis. IET Computers & Digital Techniques (CDT) pages 109-123, 2009. bib [VLSI-SOC'08] A Scratch-Pad Memory Accelerator for Exploiting Run-Time Reuse. A. Milidonis, V. Porpodas, H. Michail, A.P. Kakarountas, G. Panagiotakopoulos and CE Goutis. VLSI-SOC 2008. [DATE'07] A Decoupled Architecture of Processors with Scratch-Pad Memory Hierarchy. A. Milidonis, N. Alachiotis, V. Porpodas, H. Michail, AP Kakarountas, and CE Goutis. Design, Automation & Test in Europe (DATE), pages 1-6, 2007. bib

Links

Some of my spare-time projects:
• ZOFI Zero Overhead Fault Injection tool for super fast fault-coverage analysis.
• Xfig-slides Use Xfig to create animated slides !
• moufoplot A powerful but easy to use gnuplot front-end.
• moufodns A simple DNS server.
• How to prepare 300 USB flash drives in less than 1 Hour
• slackdesc Slackware package description generator.
• txt2bind Converts simple txt files to BIND compatible configuration files.

Other:
• Slackware
• GCC