Page Not Found
Page not found. Your pixels are in another canvas.
Sitemap
A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.
Pages
About me
About me
Posts
Can TornadoVM run Matrix Multiply faster than OpenCL Native?
Published:
This article explores how TornadoVM, a Java parallel programming framework, can outperform OpenCL code on GPUs using the Matrix Multiplication application as an example.
Using oneAPI Construction Kit and TornadoVM to accelerate Java Programs on x86, ARM and RISC-V CPUs
Published:
Running TornadoVM via the oneAPI Construction Kit for Intel, ARM and RISC-V CPUs.
Book Review: JVM Performance Engineering
Published:
Running TornadoVM on CPUs and FPGAs via oneAPI
Published:
This post shows the main steps to install and run TornadoVM on CPUs and FPGAs using the Intel oneAPI runtime for OpenCL.
Multi-device & Multi-backend TornadoVM
Published:
This post shows, via examples, how developers can benefit from these features, and reason about performance using the TornadoVM profiler to help us tune our applications.
The TornadoVM Programming Model Explained
Published:
You are a Java developer and you want to access GPUs? In this post I explain how by using TornadoVM.
Running TornadoVM within IntelliJ
Published:
Running Java applications from existing IDEs can be a cumbersome process, especially if we need to specify shared libraries. In this post, I will explain how to get access to NVIDIA and Intel-integrated GPUs from IntelliJ using TornadoVM.
Installing the NVIDIA Drivers and CUDA 12.3 on Fedora 39 with Secure Boot Enabled
Published:
Installing the NVIDIA Drivers and CUDA 12.3 on Fedora 39 with Secure Boot Enabled.
Accessible Dynamic SPIR-V Code Generation from Java
Published:
Dynamic SPIR-V Code Generation from Java. Why do we need this and how can be used?
Unified Shared Memory: Friend or Fue?
Published:
Unified Shared Memory: Friend or Fue? Understanding the Implications of Unified Memory on Managed
Configuration of the NVIDIA and Intel GPU drivers for RHEL9
Published:
This post shows the installation steps to obtain NVIDIA CUDA and Intel OpenCL and Level Zero runtimes to run applications on GPUs with RHEL.
Running TornadoVM on NVIDIA Jetson Nano
Published:
Did you know that TornadoVM can also run on ARM-based systems with NVIDIA GPUs? In this post, we will show how TornadoVM can be used on an NVIDIA Jetson Nano, a small, powerful computer designed for embedded artificial intelligence (AI) and machine learning (ML) applications.
Enabling Transparent Acceleration of Big Data Frameworks Using Heterogeneous Hardware
Published:
Exploiting heterogeneous hardware for Big Data workloads is usually done by introducing new APIs, resulting in more complex programs to develop, understand, and maintain. But, what if we do not change/extend the original programming model? Is it possible? This post discusses a new approach to do so.
Exploring Level Zero resources: repositories and purpose
Published:
Sometimes, it is not clear which Level Zero repository is the right one for our needs. In this post, we will explain each of the Level Zero public resources and what they are intended to be.
Installing CUDA, OpenCL and Level Zero in OpenSUSE Leap 15
Published:
In this post, we show how to install the NVIDIA drivers to get access to CUDA and OpenCL parallel programming frameworks and utilities for NVIDIA GPUs. We also show how to install the Intel compute-runtime drivers for accessing, via OpenCL and Level Zero, Intel Integrated Graphics.
TornadoVM Internals: Java APIs for Compiling Java methods to SPIR-V and running on GPUs via Level Zero
Published:
This post shows how to use the internal APIs to interact directly with the TornadoVM JIT compiler interface and runtime system.
Running Java Programs on XPUs with TornadoVM via Docker
Published:
In this post, we will show how to launch and accelerate Java programs on heterogeneous hardware via TornadoVM with minimal configuration using pre-built Docker images
Running TornadoVM on Intel GPUs using Windows Subsystem for Linux (WSL) for Windows 11
Published:
In this post, I will show you how we can enable TornadoVM to run on Intel HD Graphics via the OpenCL and SPIR-V Backends within WSL using Windows 11
Overall Performance of Unified Shared Memory Types with Level Zero on Intel Integrated GPUs
Published:
Does share memory really impact performance if we measure end-to-end applications on GPUs? In this post, we try to answer this question.
Understanding Memory Allocation Size Limitations with Level Zero
Published:
In this post we want to explore the memory capabilities of the Level Zero API, and, examine its constraints with respect to memory allocation.
Profiling OpenCL and SPIRV code from TornadoVM using VTune
Published:
Profiling OpenCL and SPIRV code from TornadoVM using VTune : https://jjfumero.github.io/posts/2022/02/profiling-tornadovm-with-intel-vtune/
Measuring Kernel Time and Data Transfers with Level Zero
Published:
Measuring Kernel Time and Data Transfers with Level Zero : https://jjfumero.github.io/posts/2021/09/timers-with-level-zero/
Introduction to Level Zero API for Heterogeneous Programming
Published:
Overview of the Intel Level-Zero API and a practical example to dispatch a SPIR-V kernel on the Intel HD Graphics: https://jjfumero.github.io/posts/2021/09/introduction-to-level-zero/
podcasts
portfolio
Portfolio item number 1
Published:
Short description of portfolio item number 1
Portfolio item number 2
Published:
Short description of portfolio item number 2
projects
publications
Just-In-Time GPU Compilation for Interpreted Languages with Partial Evaluation
Published in VEE 2017, Xian, China, 2017
Recommended citation: Juan Fumero, Michel Steuwer, Lukas Stadler, and Christophe Dubach. 2017. Just-In-Time GPU Compilation for Interpreted Languages with Partial Evaluation. SIGPLAN Not. 52, 7 (April 2017), 60-73. DOI: https://doi.org/10.1145/3140607.3050761 https://dl.acm.org/citation.cfm?id=3050761
PhD Thesis: Accelerating Interpreted Programming Languages on GPUs with Just-In-Time and Runtime Optimisations
Published in The University of Edinburgh, UK, 2017
Recommended citation: Juan Fumero. PhD Thesis: Accelerating Interpreted Programming Languages on GPUs with Just-In-Time and Runtime Optimisations. The University of Edinburgh, UK. August 2017. https://www.era.lib.ed.ac.uk/handle/1842/28718
Towards Practical Heterogeneous Virtual Machines
Published in MoreVMs 2018, 2018
Heterogeneous VMs, JVM, Graal, GPUs
Recommended citation: James Clarkson, Juan Fumero, Michalis Papadimitriou, Maria Xekalaki, Christos Kotselidis. Towards Practical Heterogeneous Virtual Machines. MoreVMs 2018 https://2018.programming-conference.org/track/MoreVMs-2018#program
Enabling RISC-V support on the MaxineVM
Published in RISC-V Barcelona 2018, 2018
RISC-V, JVMs, MaxineVM
Recommended citation: Foivos S Zakkak, Juan Fumero and Christos Kotselidis. Enabling RISC-V support on the MaxineVM. RISC-V Workshop, Barcelona 2018 https://www.researchgate.net/publication/325113454_Enabling_RISC-V_support_on_MaxineVM
Exploiting High-Performance Heterogeneous Hardware for Java Programs using Graal
Published in ManLang 2018, Linz, Austria, 2018
Recommended citation: James Clarkson, Juan Fumero, Michalis Papadimitriou, Foivos S. Zakkak, Maria Xekalaki, Christos Kotselidis, Mikel Lujan (The University of Manchester) Exploiting High-Performance Heterogeneous Hardware for Java Programs using Graal. ManLang 2018. https://www.researchgate.net/publication/327097904_Exploiting_High-Performance_Heterogeneous_Hardware_for_Java_Programs_using_Graal
ScootR: Scaling R Dataframes on Dataflow Systems
Published in SOCC 2018, California, 2018
Recommended citation: Andreas Kunft, Lukas Stadler, Daniele Bonetta, Cosmin Basca, Jens Meiners, Sebastian Breß, Tilmann Rabl, Juan Fumero, Volker Markl. ScootR: Scaling R Dataframes on Dataflow Systems. SOCC 2018 http://www.user.tu-berlin.de/akunft/paper/socc18-paper35.pdf
Using Compiler Snippets to Exploit Parallelism on Heterogeneous Hardware: A Java Reduction Case Study
Published in Boston, Massachusetts, United States, 2018
Recommended citation: Juan Fumero, Christos Kotselidis. Using Compiler Snippets to Exploit Parallelism on Heterogeneous Hardware: A Java Reduction Case Study. VMIL 2018 https://bit.ly/2Rh5t5c
Dynamic Application Reconfiguration on Heterogeneous Hardware
Published in VEE 2019, Providence, Rhode Island, United States, 2019
Recommended citation: Juan Fumero, Michail Papadimitriou, Foivos S. Zakkak, Maria Xekalaki, James Clarkson, and Christos Kotselidis. 2019. Dynamic application reconfiguration on heterogeneous hardware. In Proceedings of the 15th ACM SIGPLAN/SIGOPS International Conference on Virtual Execution Environments (VEE 2019). ACM, New York, NY, USA, 165-178. DOI: https://doi.org/10.1145/3313808.3313819 https://github.com/jjfumero/jjfumero.github.io/blob/master/files/VEE2019_Fumero_Preprint.pdf
Towards Prototyping and Acceleration of Java Programs onto Intel FPGAs
Published in 2019 IEEE 27th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM), 2019
Recommended citation: M. Papadimitriou, J. Fumero, A. Stratikopoulos and C. Kotselidis. Towards Prototyping and Acceleration of Java Programs onto Intel FPGAs. In 2019 IEEE 27th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM), San Diego, CA, USA, 2019 pp. 310-310. doi: 10.1109/FCCM.2019.00051 url: https://doi.ieeecomputersociety.org/10.1109/FCCM.2019.00051
Running Parallel Bytecode Interpreters on Heterogeneous Hardware
Published in MoreVMs 2020, Workshop collocated with Programming 2020. Porto, Portugal, 2020
Recommended citation: J. Fumero, A. Stratikopoulos and C. Kotselidis. Running Parallel Bytecode Interpreters on Heterogeneous Hardware. MoreVMs 2020.
Transparent Compiler and Runtime Specializations for Accelerating Managed Languages on FPGAs
Published in The Art, Science, and Engineering of Programming (Programming) 2021, 2021
Recommended citation: Michail Papadimitriou, Juan Fumero, Athanasios Stratikopoulos, Foivos Zakkak and Christos Kotselidis. Transparent Compiler and Runtime Specializations for Accelerating Managed Languages on FPGAs. Programming 2021.
Automatically Exploiting the Memory Hierarchy of GPUs through Just-in-Time Compilation
Published in VEE 2021, Virtual Execution Environments, 2021
Recommended citation: Michail Papadimitriou, Juan Fumero, Athanasios Stratikopoulos and Christos Kotselidis. Automatically Exploiting the Memory Hierarchy of GPUs through Just-in-Time Compilation . VEE 2021.
Multiple-Tasks on Multiple-Devices (MTMD): Exploiting Concurrency in Heterogeneous Managed Runtimes
Published in VEE 2021, Virtual Execution Environments, 2021
Recommended citation: Michail Papadimitriou, Eleni Markou, Juan Fumero, Athanasios Stratikopoulos, Florin Blanaru and Christos Kotselidis. Multiple-Tasks on Multiple-Devices (MTMD): Exploiting Concurrency in Heterogeneous Managed Runtimes . VEE 2021.
Enabling Pipeline Parallelism in Heterogeneous Managed Runtime Environments via Batch Processing
Published in VEE 2022, Virtual Execution Environments, 2022
Recommended citation: Florin Blanaru, Athanasios Stratikopoulos, Juan Fumero, Christos Kotselidis. VEE 2022.
Enabling Transparent Acceleration of Big Data Frameworks Using Heterogeneous Hardware
Published in VLDB 2023, 49th International Conference on Very Large Data Bases, 2022
Recommended citation: Maria Xekalaki, Juan Fumero, Athanasios Stratikopoulos, Katerina Doka, Christos, Katsakioris, Constantinos Bitsakos, Nectarios Koziris, Christos Kotselidis. VLDB23
Cross-Language Interoperability of Heterogeneous Code
Published in MoreVMs 2023 - Collocated with Programming, 2023
Recommended citation: Athanasios Stratikopoulos, Florin Blanaru, Juan Fumero, Maria Xekalaki, Orion Papadakis, Christos Kotselidis. MoreVMs 2023
Experiences in Building a Composable and Functional API for Runtime SPIR-V Code Generation
Published in Preprint, 2023
Recommended citation: Fumero, J., Rethy, G., Stratikopoulos, A., Foutris, N., & Kotselidis, C. (2023). Experiences in Building a Composable and Functional API for Runtime SPIR-V Code Generation. ArXiv. /abs/2305.09493
Unified Shared Memory: Friend or Foe? Understanding the Implications of Unified Memory on Managed Heaps
Published in MPLR 2023, 2023
Recommended citation: Juan Fumero, Florin Blanaru, Athanasios Stratikopoulos, Steve Dohrmann, Sandhya Viswanathan, Christos Kotselidis. Unified Shared Memory: Friend or Foe? Understanding the Implications of Unified Memory on Managed Heaps. MPLR 2023.
Beehive SPIR-V Toolkit: A Composable and Functional API for Runtime SPIR-V Code Generation
Published in VMIL 2023, 2023
Recommended citation: Fumero, J., Rethy, G., Stratikopoulos, A., Foutris, N., & Kotselidis, C. Beehive SPIR-V Toolkit: A Composable and Functional API for Runtime SPIR-V Code Generation. (VMIL 2023).
[BOOK] Programming Heterogeneous Hardware via Managed Runtime Systems
Published in SpringerBriefs in Computer Science, 2024
SpringerBriefs in Computer Science
Recommended citation: Juan Fumero, Athanasios Stratikopoulos, Christos Kotselidis. Programming Heterogeneous Hardware via Managed Runtime Systems. SpringerBriefs in Computer Science 2024. https://link.springer.com/book/10.1007/978-3-031-49559-5
service
supervision
talks
Invited Talk - FastR-Flink: A compiler based approach for distributed computing in R
Published:
During the past few years R has become an important language for data analysis, data representation and visualization. R is a very expressive language which combines functional and dynamic aspects, with laziness and object oriented programming. However, the default Rimplementation is neither fast nor distributed, both features crucial for “big data” processing.
Invited Talk at Edinburgh University - FastR-Flink
Published:
During the past few years R has become an important language for data analysis, data representation and visualization. R is a very expressive language which combines functional and dynamic aspects, with laziness and object oriented programming. However, the default Rimplementation is neither fast nor distributed, both features crucial for “big data” processing.
Invited Talk - OpenCL Just-In-Time Compilation for Dynamic Programming Languages
Published:
In this talk we present a technique to automatically offload parts of the input program written in a dynamic language into OpenCL without any changes in the original source code. Our preliminary results show we achieve speedups of up to 150x when using the GPU (3) OpenCL JIT Compilation for Dynamic Programming Languages.
Towards Practical Heterogeneous Virtual Machines
Published:
Heterogeneous computing emerged as a means to achieve higher performance and energy efficiency. However, this trend has been accompanied by changes in software development norms that do not necessarily favour programmers. A prime example is the two most popular heterogeneous programming languages, CUDA and OpenCL, which expose several low-level features to the API making them difficult to use by non-expert users.
Invited talk at MSR - Tornado VM: A Virtual Machine for Exploiting High-Performance Heterogeneous Hardware of Java Programs
Published:
The proliferation of heterogeneous hardware in recent years means that every system we program is likely to include a mix of computing elements; each of these with different characteristics. This trend has been accompanied by changes in software development norms that do not necessarily favor programmers. A prime example is the two most popular heterogeneous programming languages, CUDA and OpenCL, which expose several low-level features to the API making them difficult to use by non-expert users.
Invited talk at ARM - Exploiting Heterogeneous Hardware from Managed Runtime Languages
Published:
The proliferation of heterogeneous hardware in recent years means that every system we program is likely to include a mix of computing elements; each of these with different characteristics. This trend has been accompanied by changes in software development norms that do not necessarily favor programmers. A prime example is the two most popular heterogeneous programming languages, CUDA and OpenCL, which expose several low-level features to the API making them difficult to use by non-expert users.
TornadoVM: A Virtual Machine for Exploiting High-Performance Heterogeneous Hardware of Java Programs
Published:
Talk about TornadoVM given at the JVMLS 2019 workshop.
TornadoVM: A virtual machine for exploiting high performance heterogeneous hardware
Published:
Talk about TornadoVM given at Joker<?> 2019 Conference.
TornadoVM: Java for GPUs and FPGAs @QCon-London
Published:
Slides available here
Transparent Heterogeneous Computing for Java via TornadoVM @ NYJavaSIG
Published:
Slides available here
Designing Parallel Programming APIs for Heterogeneous Hardware on top of Managed Runtime Systems
Published:
teaching
Object Oriented Programming (Java course)
Tutoring, The University of Edinburgh, 2015
I was tutoring the Introduction to Object Obriented Programming through Java to groups of 8-10 of undergraduate students during Spring 2015 and 2016. We covered an introduction to topics such as Object Oriented Programming, inheritance, Java collections, Threads, concurrency and parallelism. We also covered some material about how the Java Virtual Machine works.