Talks | Parallel Computing
2024
Gianluca Mittone, Alberto Mulone, Iacopo Colonnelli, Robert Birke, Marco Aldinucci
Enabling Cross-Facility LLMs Pre-Training Miscellaneous
Accelerating the Development and Use of Generative AI for Science and Engineering: The Trillion Parameter Consortium (TPC), 2024.
Abstract | Links | BibTeX | Tags: eupilot, fl, icsc, space, streamflow
@misc{24:mittone:TPC,
title = {Enabling Cross-Facility LLMs Pre-Training},
author = {Gianluca Mittone and Alberto Mulone and Iacopo Colonnelli and Robert Birke and Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/DRgm8ebBkKQgD2d},
year = {2024},
date = {2024-11-01},
address = {Atlanta, GE, USA},
abstract = {Big-tech companies pre-train SOTA LLMs on special-purpose, private HPCs, while public research centres lack the resources to compete. We advocate a new take on large model training, e.g., LLMs, called xFFL, which leverages federated learning as an enabling technique to exploit geographically distributed computing power to bridge such digital divide. This work introduces a proof-of-concept federated training of LLaMA-3 8B on three EuroHPC Top500 facilities, proving the viability of leveraging cross-facility publicly available computational power to sustain SOTA LLM workloads.},
howpublished = {Accelerating the Development and Use of Generative AI for Science and Engineering: The Trillion Parameter Consortium (TPC)},
keywords = {eupilot, fl, icsc, space, streamflow},
pubstate = {published},
tppubtype = {misc}
}
Big-tech companies pre-train SOTA LLMs on special-purpose, private HPCs, while public research centres lack the resources to compete. We advocate a new take on large model training, e.g., LLMs, called xFFL, which leverages federated learning as an enabling technique to exploit geographically distributed computing power to bridge such digital divide. This work introduces a proof-of-concept federated training of LLaMA-3 8B on three EuroHPC Top500 facilities, proving the viability of leveraging cross-facility publicly available computational power to sustain SOTA LLM workloads.
Gianluca Mittone
Benchmarking HPC Performance for State-of-the-Art AI Workloads Miscellaneous
2024.
Abstract | Links | BibTeX | Tags: ai, eupilot, icsc
@misc{24:mittone:itadata:shpcpee,
title = {Benchmarking HPC Performance for State-of-the-Art AI Workloads},
author = {Gianluca Mittone},
url = {https://datacloud.di.unito.it/index.php/s/5Ep3W7cPW5baZfr},
year = {2024},
date = {2024-09-01},
address = {Pisa, Italy},
abstract = {Benchmarking the performance of modern High-Performance Computing (HPC) infrastructure on Artificial Intelligence (AI) workloads is a hot topic in the supercomputing community. While research communities and big-tech companies actively invest in larger, more powerful data centres to support AI research, the standard computational performance benchmarking tools (e.g., LINPACK) are increasingly becoming outdated since they are not specifically tailored for AI workloads. Some tools, such as MLPerf, are trying to bridge this gap, but the HPC community still has not adopted them as standards. Since this trend became particularly evident with the advent of Large Language Models (LLMs), this work will delve into LLM training at scale as a way to benchmark Top500 HPC infrastructures on current AI workloads. The scalability performances of a major LLM model (i.e., Meta's LLaMA) on different HPCs (Leonardo, LUMI, MeluXina, Karolina) are exposed and discussed along with their Top500 positioning.
However, it should be noted that state-of-the-art LLM models are not trained on thousands of computing nodes but on hundreds. This choice is due to multiple factors, such as the influence of the training scaling on the model's convergence and the instability of large-scale deployments due to hardware/software failure. A benchmarking approach based on the next-generation LLM training approach is proposed to bypass all these issues. State-of-the-art LLMs are not monolithic structures but Mixture-of-Experts (MoE) models; this design implies innovative frontiers for the distributed training of such models due to the experts' training being potentially more parallelisable than a single monolithic model. We thus propose to create an AI-oriented HPC benchmark suite based on the parallel training of MoE models to measure the throughput performance of HPC systems on state-of-the-art AI workloads.},
keywords = {ai, eupilot, icsc},
pubstate = {published},
tppubtype = {misc}
}
Benchmarking the performance of modern High-Performance Computing (HPC) infrastructure on Artificial Intelligence (AI) workloads is a hot topic in the supercomputing community. While research communities and big-tech companies actively invest in larger, more powerful data centres to support AI research, the standard computational performance benchmarking tools (e.g., LINPACK) are increasingly becoming outdated since they are not specifically tailored for AI workloads. Some tools, such as MLPerf, are trying to bridge this gap, but the HPC community still has not adopted them as standards. Since this trend became particularly evident with the advent of Large Language Models (LLMs), this work will delve into LLM training at scale as a way to benchmark Top500 HPC infrastructures on current AI workloads. The scalability performances of a major LLM model (i.e., Meta's LLaMA) on different HPCs (Leonardo, LUMI, MeluXina, Karolina) are exposed and discussed along with their Top500 positioning.
However, it should be noted that state-of-the-art LLM models are not trained on thousands of computing nodes but on hundreds. This choice is due to multiple factors, such as the influence of the training scaling on the model's convergence and the instability of large-scale deployments due to hardware/software failure. A benchmarking approach based on the next-generation LLM training approach is proposed to bypass all these issues. State-of-the-art LLMs are not monolithic structures but Mixture-of-Experts (MoE) models; this design implies innovative frontiers for the distributed training of such models due to the experts' training being potentially more parallelisable than a single monolithic model. We thus propose to create an AI-oriented HPC benchmark suite based on the parallel training of MoE models to measure the throughput performance of HPC systems on state-of-the-art AI workloads.
However, it should be noted that state-of-the-art LLM models are not trained on thousands of computing nodes but on hundreds. This choice is due to multiple factors, such as the influence of the training scaling on the model's convergence and the instability of large-scale deployments due to hardware/software failure. A benchmarking approach based on the next-generation LLM training approach is proposed to bypass all these issues. State-of-the-art LLMs are not monolithic structures but Mixture-of-Experts (MoE) models; this design implies innovative frontiers for the distributed training of such models due to the experts' training being potentially more parallelisable than a single monolithic model. We thus propose to create an AI-oriented HPC benchmark suite based on the parallel training of MoE models to measure the throughput performance of HPC systems on state-of-the-art AI workloads.
Gianluca Mittone
Pushing Federated Learning Boundaries: Three Innovative Distributed Intelligence Approaches Miscellaneous
2024.
Abstract | Links | BibTeX | Tags: eupilot, fl, icsc
@misc{24:mittone:bighpc,
title = {Pushing Federated Learning Boundaries: Three Innovative Distributed Intelligence Approaches},
author = {Gianluca Mittone},
url = {https://datacloud.di.unito.it/index.php/s/eKbRtSAEdmSFJYW},
year = {2024},
date = {2024-09-01},
address = {Pisa, Italy},
abstract = {Federated learning is a distributed, privacy-preserving machine learning technique used on private, decentralised data. It allows multiple parties to cooperatively solve a common machine learning problem without sharing the local data. Three assumptions of state-of-the-art federated learning software constitute the starting points for this research work: 1) their inner workings being strictly tied to deep learning models, 2) the centralised structure currently implemented by many commercial frameworks, and 3) their assumption of being deployed on private, specialised computing infrastructures. The proposed research expands the federated learning paradigm to handle scenarios in which these three conditions do not hold. Such research problems are addressed methodologically and practically, and three open-source, proof-of-concept software are made freely available as tangible research results: OpenFL-x, FastFL, and xFFL.},
keywords = {eupilot, fl, icsc},
pubstate = {published},
tppubtype = {misc}
}
Federated learning is a distributed, privacy-preserving machine learning technique used on private, decentralised data. It allows multiple parties to cooperatively solve a common machine learning problem without sharing the local data. Three assumptions of state-of-the-art federated learning software constitute the starting points for this research work: 1) their inner workings being strictly tied to deep learning models, 2) the centralised structure currently implemented by many commercial frameworks, and 3) their assumption of being deployed on private, specialised computing infrastructures. The proposed research expands the federated learning paradigm to handle scenarios in which these three conditions do not hold. Such research problems are addressed methodologically and practically, and three open-source, proof-of-concept software are made freely available as tangible research results: OpenFL-x, FastFL, and xFFL.
Giulio Malenza
Preliminary analysis of model parallelism applications on a 64-core RV64 Server CPU Miscellaneous
2024.
Abstract | Links | BibTeX | Tags: eupilot, icsc
@misc{24:gmalenza:hlpp:MPRISC-v,
title = {Preliminary analysis of model parallelism applications on a 64-core RV64 Server CPU},
author = {Giulio Malenza},
url = {https://datacloud.di.unito.it/index.php/s/JrWwKALeaFEJSQo},
year = {2024},
date = {2024-07-01},
address = {Pisa, Italy},
abstract = {Massive Data Parallel workloads, driven by inference on large ML models, are pushing hardware vendors to develop efficient and cost-effective multi-core server CPUs. The RISC-V architecture plays a prominent role due to its open, extensible and energy-friendly ISA. Despite significant progress in recent years, finding efficient methods to run parallel applications on new architectures to harness their maximum performance fully remains a challenge. In this study, we benchmark the inference of machine learning models on the SOPHON SG2042 SoC, the first server-grade CPU based on the RV64 ISA, composed of 64 cores arranged in a grid of 16 groups of 4 cores. Specifically, we aim to enhance performance via better cache hit ratios stemming from model parallelism to split and assign parts of the model to specific (groups of) cores using a pipeline execution. We orchestrate execution using FastFlow, a low-level programming framework designed for multithreaded streaming applications. By comparing the results against the standard multi-core inference and analyzing the effects of different submodel-to-core mapping strategies, we aim to provide a comprehensive understanding of how the model parallel approach can maximize efficiency and utilization of hardware resources.},
keywords = {eupilot, icsc},
pubstate = {published},
tppubtype = {misc}
}
Massive Data Parallel workloads, driven by inference on large ML models, are pushing hardware vendors to develop efficient and cost-effective multi-core server CPUs. The RISC-V architecture plays a prominent role due to its open, extensible and energy-friendly ISA. Despite significant progress in recent years, finding efficient methods to run parallel applications on new architectures to harness their maximum performance fully remains a challenge. In this study, we benchmark the inference of machine learning models on the SOPHON SG2042 SoC, the first server-grade CPU based on the RV64 ISA, composed of 64 cores arranged in a grid of 16 groups of 4 cores. Specifically, we aim to enhance performance via better cache hit ratios stemming from model parallelism to split and assign parts of the model to specific (groups of) cores using a pipeline execution. We orchestrate execution using FastFlow, a low-level programming framework designed for multithreaded streaming applications. By comparing the results against the standard multi-core inference and analyzing the effects of different submodel-to-core mapping strategies, we aim to provide a comprehensive understanding of how the model parallel approach can maximize efficiency and utilization of hardware resources.
Robert Birke
The impact of the advances in generative models on applications and systems Miscellaneous
8th GDR RSD / ASF Winter School on Distributed Systems & Networks 2024, 2024, (Keynote talk).
Abstract | Links | BibTeX | Tags: ai, eupilot, textarossa
@misc{24:ASF:WINTER,
title = {The impact of the advances in generative models on applications and systems},
author = {Robert Birke},
url = {https://datacloud.di.unito.it/index.php/s/QYTCMfWp4sY5qx4},
year = {2024},
date = {2024-01-01},
address = {Le Pleynet, France},
abstract = {Generative models have achieved unprecedented quality levels across a wide range of data types. This advance often stems from the ever increasing data and compute used to train larger and larger models. One major use case of such synthetic data is in privacy-compliant data sharing. Gartner predicts that synthetic data will reduce by 2025 the need for real data by 70% for analytics and machine learning. We will look at generative models, with a special focus on tabular data, and the issue of democratization of large model training.},
howpublished = {8th GDR RSD / ASF Winter School on Distributed Systems & Networks 2024},
note = {Keynote talk},
keywords = {ai, eupilot, textarossa},
pubstate = {published},
tppubtype = {misc}
}
Generative models have achieved unprecedented quality levels across a wide range of data types. This advance often stems from the ever increasing data and compute used to train larger and larger models. One major use case of such synthetic data is in privacy-compliant data sharing. Gartner predicts that synthetic data will reduce by 2025 the need for real data by 70% for analytics and machine learning. We will look at generative models, with a special focus on tabular data, and the issue of democratization of large model training.
Gianluca Mittone
RISC-V for AI Miscellaneous
High Performance, Edge And Cloud computing Conference 2024 (HiPEAC 2024), 2024.
Abstract | Links | BibTeX | Tags: eupilot, icsc
@misc{24:HiPEAC:riscv,
title = {RISC-V for AI},
author = {Gianluca Mittone},
url = {https://datacloud.di.unito.it/index.php/s/rFtxT7zryoKNGbP},
year = {2024},
date = {2024-01-01},
address = {Garching bei München, München, Germany},
abstract = {AI-focused RISC-V-based hardware accelerators},
howpublished = {High Performance, Edge And Cloud computing Conference 2024 (HiPEAC 2024)},
keywords = {eupilot, icsc},
pubstate = {published},
tppubtype = {misc}
}
AI-focused RISC-V-based hardware accelerators
2023
Marco Aldinucci
Federated Learning: A Distributed System Viewpoint Miscellaneous
Bicocca University seminars, Milan, Italy, 2023, (Invited talk).
Abstract | Links | BibTeX | Tags: eupilot, icsc, textarossa
@misc{23:FL:bicocca,
title = {Federated Learning: A Distributed System Viewpoint},
author = {Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/FfEzADQtC73GgLs},
year = {2023},
date = {2023-12-01},
abstract = {Decentralized machine learning (DML) enables collaborative machine learning without centralized input data. Federated learning (FL) and edge inference (EI) are examples of DML. Collaboration naturally happens at the edge of a distributed system with inherently distributed data. While tools for DML are starting to flourish, much needs to be done to get more flexible and portable tools to experiment with novel techniques, non-fully connected topologies, multiple data domains, and asynchronous collaboration schemes. We'll present recent advances in DML, aiming to improve usability in data centers and, at the edge, to widen the class of models extending FL to non-DDN paradigms, to improve the accuracy of models controlling normalization and frequency of communications, and to boost data privacy through generative adversarial networks.},
howpublished = {Bicocca University seminars, Milan, Italy},
note = {Invited talk},
keywords = {eupilot, icsc, textarossa},
pubstate = {published},
tppubtype = {misc}
}
Decentralized machine learning (DML) enables collaborative machine learning without centralized input data. Federated learning (FL) and edge inference (EI) are examples of DML. Collaboration naturally happens at the edge of a distributed system with inherently distributed data. While tools for DML are starting to flourish, much needs to be done to get more flexible and portable tools to experiment with novel techniques, non-fully connected topologies, multiple data domains, and asynchronous collaboration schemes. We'll present recent advances in DML, aiming to improve usability in data centers and, at the edge, to widen the class of models extending FL to non-DDN paradigms, to improve the accuracy of models controlling normalization and frequency of communications, and to boost data privacy through generative adversarial networks.
Iacopo Colonnelli, Robert Birke, Giulio Malenza, Gianluca Mittone, Alberto Mulone, Marco Aldinucci, Valerio Basile, Marco Antonio Stranisci, Viviana Patti, Jeroen Galjaard, Lydia Y. Chen, Sanzio Bassini, Massimiliano Guarrasi, Gabriella Scipione, Jan Martinovič, Vit Vondrák
Cross-Facility Federated Learning Miscellaneous
1st EuroHPC User Day, 2023.
Links | BibTeX | Tags: across, ai, eupex, eupilot, HPC
@misc{23:eurohpc,
title = {Cross-Facility Federated Learning},
author = {Iacopo Colonnelli and Robert Birke and Giulio Malenza and Gianluca Mittone and Alberto Mulone and Marco Aldinucci and Valerio Basile and Marco Antonio Stranisci and Viviana Patti and Jeroen Galjaard and Lydia Y. Chen and Sanzio Bassini and Massimiliano Guarrasi and Gabriella Scipione and Jan Martinovič and Vit Vondrák},
url = {https://datacloud.di.unito.it/index.php/s/DDAz4QkJP3WZ68M},
year = {2023},
date = {2023-12-01},
address = {Bruxelles, Belgium},
howpublished = {1st EuroHPC User Day},
keywords = {across, ai, eupex, eupilot, HPC},
pubstate = {published},
tppubtype = {misc}
}
Gianluca Mittone, Giulio Malenza, Marco Aldinucci, Robert Birke
Distributed Edge Inference: an Experimental Study on Multiview Detection Miscellaneous
The 16th IEEE/ACM International Conference on Utility and Cloud Computing (UCC 2023), 2023.
Abstract | Links | BibTeX | Tags: ai, eupilot, icsc
@misc{23:ucc:multiview,
title = {Distributed Edge Inference: an Experimental Study on Multiview Detection},
author = {Gianluca Mittone and Giulio Malenza and Marco Aldinucci and Robert Birke},
url = {https://datacloud.di.unito.it/index.php/s/XfjNZEPSNfSKPFr},
year = {2023},
date = {2023-12-01},
address = {Taormina, Italy},
abstract = {Computing is evolving rapidly to cater to the increasing demand for sophisticated services, and Cloud computing lays a solid foundation for flexible on-demand provisioning. However, as the size of applications grows, the centralised client-server approach used by Cloud computing increasingly limits the applications scalability. To achieve ultra-scalability, cloud/edge/fog computing converges into the compute continuum, completely decentralising the infrastructure to encompass universal, pervasive resources. The compute continuum makes devising applications benefitting from this complex environment a challenging research problem. We put the opportunities the compute continuum others to the test through a real-world multi-view detection model (MvDet) implemented with the FastFL C/C++ high-performance edge inference framework. Computational performance is discussed considering many experimental scenarios, encompassing different edge computational capabilities and network bandwidths. We obtain up to 1.92x speedup in inference time over a centralised solution using the same devices.},
howpublished = {The 16th IEEE/ACM International Conference on Utility and Cloud Computing (UCC 2023)},
keywords = {ai, eupilot, icsc},
pubstate = {published},
tppubtype = {misc}
}
Computing is evolving rapidly to cater to the increasing demand for sophisticated services, and Cloud computing lays a solid foundation for flexible on-demand provisioning. However, as the size of applications grows, the centralised client-server approach used by Cloud computing increasingly limits the applications scalability. To achieve ultra-scalability, cloud/edge/fog computing converges into the compute continuum, completely decentralising the infrastructure to encompass universal, pervasive resources. The compute continuum makes devising applications benefitting from this complex environment a challenging research problem. We put the opportunities the compute continuum others to the test through a real-world multi-view detection model (MvDet) implemented with the FastFL C/C++ high-performance edge inference framework. Computational performance is discussed considering many experimental scenarios, encompassing different edge computational capabilities and network bandwidths. We obtain up to 1.92x speedup in inference time over a centralised solution using the same devices.
Samuele Fonio
Benchmarking Federated Learning Frameworks for Medical Imaging Tasks Miscellaneous
Image Analysis and Processing - ICIAP 2023 - 22th International Conference - FedMed, 2023.
Abstract | Links | BibTeX | Tags: ai, eupilot, fl, icsc
@misc{23:iciap:benchmed,
title = {Benchmarking Federated Learning Frameworks for Medical Imaging Tasks},
author = {Samuele Fonio},
url = {https://datacloud.di.unito.it/index.php/s/sR7YeTGgfH4DtCR},
year = {2023},
date = {2023-09-01},
address = {Udine, Italy},
abstract = {This paper presents a comprehensive benchmarking study of various Federated Learning (FL) frameworks applied to the task of Medical Image Classification. The research specifically addresses the often neglected and complex aspects of scalability and usability in off-the-shelf FL frameworks. Through experimental validation using real case deployments, we provide empirical evidence of the performance and practical relevance of open source FL frameworks. Our findings contribute valuable insights for anyone interested in deploying a FL system, with a particular focus on the healthcare domain—an increasingly attractive field for FL applications.},
howpublished = {Image Analysis and Processing - ICIAP 2023 - 22th International Conference - FedMed},
keywords = {ai, eupilot, fl, icsc},
pubstate = {published},
tppubtype = {misc}
}
This paper presents a comprehensive benchmarking study of various Federated Learning (FL) frameworks applied to the task of Medical Image Classification. The research specifically addresses the often neglected and complex aspects of scalability and usability in off-the-shelf FL frameworks. Through experimental validation using real case deployments, we provide empirical evidence of the performance and practical relevance of open source FL frameworks. Our findings contribute valuable insights for anyone interested in deploying a FL system, with a particular focus on the healthcare domain—an increasingly attractive field for FL applications.
Gianluca Mittone, Samuele Fonio
Benchmarking Federated Learning Scalability Miscellaneous
2nd Italian Conference on Big Data and Data Science (ITADATA 2023), 2023.
Abstract | Links | BibTeX | Tags: ai, eupilot, fl, icsc
@misc{23:itadata:fl_scaling,
title = {Benchmarking Federated Learning Scalability},
author = {Gianluca Mittone and Samuele Fonio},
url = {https://datacloud.di.unito.it/index.php/s/QZGxC4X3s5LG5oT},
year = {2023},
date = {2023-09-01},
address = {Naples, Italy},
abstract = {Federated Learning (FL) is a widespread Machine Learning paradigm handling distributed Big Data. In this work, we demonstrate that different FL frameworks expose different scaling performances despite adopting the same technologies, highlighting the need for a more comprehensive study on the topic.},
howpublished = {2nd Italian Conference on Big Data and Data Science (ITADATA 2023)},
keywords = {ai, eupilot, fl, icsc},
pubstate = {published},
tppubtype = {misc}
}
Federated Learning (FL) is a widespread Machine Learning paradigm handling distributed Big Data. In this work, we demonstrate that different FL frameworks expose different scaling performances despite adopting the same technologies, highlighting the need for a more comprehensive study on the topic.
Gianluca Mittone, Walter Riviera, Iacopo Colonnelli, Robert Birke, Marco Aldinucci
Model-Agnostic Federated Learning Miscellaneous
29th International European Conference on Parallel and Distributed Computing (Euro-Par '23), 2023.
Abstract | Links | BibTeX | Tags: ai, eupilot, icsc
@misc{23:europar:mafl,
title = {Model-Agnostic Federated Learning},
author = {Gianluca Mittone and Walter Riviera and Iacopo Colonnelli and Robert Birke and Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/9T6G2tRreRomBAE},
year = {2023},
date = {2023-09-01},
address = {Limassol, Cyprus},
abstract = {Since its debut in 2016, Federated Learning (FL) has been tied to the inner workings of Deep Neural Networks (DNNs); this allowed its development as DNNs proliferated but neglected those scenarios in which using DNNs is not possible or advantageous. The fact that most current FL frameworks only support DNNs reinforces this problem. To address the lack of non-DNN-based FL solutions, we propose MAFL (Model-Agnostic Federated Learning). MAFL merges a model-agnostic FL algorithm, AdaBoost.F, with an open industry-grade FL framework: Intel® OpenFL. MAFL is the first FL system not tied to any machine learning model, allowing exploration of FL beyond DNNs. We test MAFL from multiple points of view, assessing its correctness, flexibility, and scaling properties up to 64 nodes of an HPC cluster. We also show how we optimised OpenFL achieving a 5.5x speedup over a standard FL scenario. MAFL is compatible with x86-64, ARM-v8, Power and RISC-V.},
howpublished = {29th International European Conference on Parallel and Distributed Computing (Euro-Par '23)},
keywords = {ai, eupilot, icsc},
pubstate = {published},
tppubtype = {misc}
}
Since its debut in 2016, Federated Learning (FL) has been tied to the inner workings of Deep Neural Networks (DNNs); this allowed its development as DNNs proliferated but neglected those scenarios in which using DNNs is not possible or advantageous. The fact that most current FL frameworks only support DNNs reinforces this problem. To address the lack of non-DNN-based FL solutions, we propose MAFL (Model-Agnostic Federated Learning). MAFL merges a model-agnostic FL algorithm, AdaBoost.F, with an open industry-grade FL framework: Intel® OpenFL. MAFL is the first FL system not tied to any machine learning model, allowing exploration of FL beyond DNNs. We test MAFL from multiple points of view, assessing its correctness, flexibility, and scaling properties up to 64 nodes of an HPC cluster. We also show how we optimised OpenFL achieving a 5.5x speedup over a standard FL scenario. MAFL is compatible with x86-64, ARM-v8, Power and RISC-V.
Gianluca Mittone, Robert Birke, Marco Aldinucci
Model-Agnostic Federated Learning Miscellaneous
29th International European Conference on Parallel and Distributed Computing (Euro-Par '23), 2023.
Abstract | Links | BibTeX | Tags: eupilot, icsc
@misc{23:europar:phdtalk,
title = {Model-Agnostic Federated Learning},
author = {Gianluca Mittone and Robert Birke and Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/pT3qxkwzzsHR3nS},
year = {2023},
date = {2023-08-01},
address = {Limassol, Cyprus},
abstract = {Since its debut in 2016, Federated Learning (FL) has been tied to the inner workings of Deep Neural Networks (DNNs); this allowed its development as DNNs proliferated but neglected those scenarios in which using DNNs is not possible or advantageous. The fact that most current FL frameworks only support DNNs reinforces this problem. To address the lack of non-DNN-based FL solutions, we propose MAFL (Model-Agnostic Federated Learning). Decentralised Machine Learning (DML) enables collaborative machine learning without centralised input data. Federated Learning (FL) and Edge Inference are examples of DML. While tools for DML (especially FL) are starting to flourish, many are not flexible and portable enough to experiment with novel processors (e.g., RISC-V), non-fully connected network topologies, and asynchronous collaboration schemes. We overcome these limitations via a domain-specific language allowing us to map DML schemes to an underlying middleware, i.e. the FastFlow parallel programming library.},
howpublished = {29th International European Conference on Parallel and Distributed Computing (Euro-Par '23)},
keywords = {eupilot, icsc},
pubstate = {published},
tppubtype = {misc}
}
Since its debut in 2016, Federated Learning (FL) has been tied to the inner workings of Deep Neural Networks (DNNs); this allowed its development as DNNs proliferated but neglected those scenarios in which using DNNs is not possible or advantageous. The fact that most current FL frameworks only support DNNs reinforces this problem. To address the lack of non-DNN-based FL solutions, we propose MAFL (Model-Agnostic Federated Learning). Decentralised Machine Learning (DML) enables collaborative machine learning without centralised input data. Federated Learning (FL) and Edge Inference are examples of DML. While tools for DML (especially FL) are starting to flourish, many are not flexible and portable enough to experiment with novel processors (e.g., RISC-V), non-fully connected network topologies, and asynchronous collaboration schemes. We overcome these limitations via a domain-specific language allowing us to map DML schemes to an underlying middleware, i.e. the FastFlow parallel programming library.
Gianluca Mittone, Nicolò Tonci, Robert Birke, Iacopo Colonnelli, Doriana Medić, Andrea Bartolini, Roberto Esposito, Emanuele Parisi, Francesco Beneventi, Mirko Polato, Massimo Torquati, Luca Benini, Marco Aldinucci
Experimenting with Emerging RISC-V Systems for Decentralised Machine Learning Miscellaneous
20th ACM international conference on computing frontiers (CF '23), 2023, (Invited talk).
Abstract | Links | BibTeX | Tags: ai, eupilot, icsc
@misc{23:ACMCF,
title = {Experimenting with Emerging RISC-V Systems for Decentralised Machine Learning},
author = {Gianluca Mittone and Nicolò Tonci and Robert Birke and Iacopo Colonnelli and Doriana Medić and Andrea Bartolini and Roberto Esposito and Emanuele Parisi and Francesco Beneventi and Mirko Polato and Massimo Torquati and Luca Benini and Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/BYyqZbHzzN4DL8Z},
year = {2023},
date = {2023-05-01},
abstract = {Decentralised Machine Learning (DML) enables collaborative machine learning without centralised input data. Federated Learning (FL) and Edge Inference are examples of DML. While tools for DML (especially FL) are starting to flourish, many are not flexible and portable enough to experiment with novel processors (e.g., RISC-V), non-fully connected network topologies, and asynchronous collaboration schemes. We overcome these limitations via a domain-specific language allowing us to map DML schemes to an underlying middleware, i.e. the FastFlow parallel programming library. We experiment with it by generating different working DML schemes on x86-64 and ARM platforms and an emerging RISC-V one. We characterise the performance and energy efficiency of the presented schemes and systems. As a byproduct, we introduce a RISC-V porting of the PyTorch framework, the first publicly available to our knowledge.},
howpublished = {20th ACM international conference on computing frontiers (CF '23)},
note = {Invited talk},
keywords = {ai, eupilot, icsc},
pubstate = {published},
tppubtype = {misc}
}
Decentralised Machine Learning (DML) enables collaborative machine learning without centralised input data. Federated Learning (FL) and Edge Inference are examples of DML. While tools for DML (especially FL) are starting to flourish, many are not flexible and portable enough to experiment with novel processors (e.g., RISC-V), non-fully connected network topologies, and asynchronous collaboration schemes. We overcome these limitations via a domain-specific language allowing us to map DML schemes to an underlying middleware, i.e. the FastFlow parallel programming library. We experiment with it by generating different working DML schemes on x86-64 and ARM platforms and an emerging RISC-V one. We characterise the performance and energy efficiency of the presented schemes and systems. As a byproduct, we introduce a RISC-V porting of the PyTorch framework, the first publicly available to our knowledge.
Gianluca Mittone, Filip Svoboda, Marco Aldinucci, Nicholas D. Lane, Pietro Lio'
A Federated Learning Benchmark for Drug-Target Interaction Miscellaneous
2023 ACM international Web Conference (WWW '23), 2023, (Invited talk).
Abstract | Links | BibTeX | Tags: eupilot, icsc
@misc{23:WWW,
title = {A Federated Learning Benchmark for Drug-Target Interaction},
author = {Gianluca Mittone and Filip Svoboda and Marco Aldinucci and Nicholas D. Lane and Pietro Lio'},
url = {https://datacloud.di.unito.it/index.php/s/js7go3EorZxSLn9},
year = {2023},
date = {2023-05-01},
abstract = {Aggregating pharmaceutical data in the drug-target interaction (DTI) domain can potentially deliver life-saving breakthroughs. It is, however, notoriously difficult due to regulatory constraints and commercial interests. This work proposes the application of federated learning, which is reconcilable with the industry's constraints. It does not require sharing any information that would reveal the entities' data or any other high-level summary. When used on a representative GraphDTA model and the KIBA dataset, it achieves up to 15% improved performance relative to the best available non-privacy preserving alternative. Our extensive battery of experiments shows that, unlike in other domains, the non-IID data distribution in the DTI datasets does not deteriorate FL performance. Additionally, we identify a material trade-off between the benefits of adding new data and the cost of adding more clients.},
howpublished = {2023 ACM international Web Conference (WWW '23)},
note = {Invited talk},
keywords = {eupilot, icsc},
pubstate = {published},
tppubtype = {misc}
}
Aggregating pharmaceutical data in the drug-target interaction (DTI) domain can potentially deliver life-saving breakthroughs. It is, however, notoriously difficult due to regulatory constraints and commercial interests. This work proposes the application of federated learning, which is reconcilable with the industry's constraints. It does not require sharing any information that would reveal the entities' data or any other high-level summary. When used on a representative GraphDTA model and the KIBA dataset, it achieves up to 15% improved performance relative to the best available non-privacy preserving alternative. Our extensive battery of experiments shows that, unlike in other domains, the non-IID data distribution in the DTI datasets does not deteriorate FL performance. Additionally, we identify a material trade-off between the benefits of adding new data and the cost of adding more clients.
Marco Aldinucci
Experimenting with Systems for Decentralized Machine Learning Miscellaneous
NVidia GTC 2023, 2023.
Abstract | Links | BibTeX | Tags: across, admire, epi, eumaster4hpc, eupex, eupilot, hpc4ai, space, textarossa
@misc{23:gtc:fl,
title = {Experimenting with Systems for Decentralized Machine Learning},
author = {Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/oyLt7xwkbKxz65c},
year = {2023},
date = {2023-03-01},
abstract = {Decentralized machine learning (DML) enables collaborative machine learning without centralized input data. Federated learning (FL) and edge inference (EI) are examples of DML. Collaboration naturally happens at the edge of a distributed system with inherently distributed data. While tools for DML are starting to flourish, much needs to be done to get more flexible and portable tools to experiment with novel techniques, non-fully connected topologies, multiple data domains, and asynchronous collaboration schemes. We'll present recent advances in DML, aiming to improve usability in data centers and, at the edge, to widen the class of models extending FL to non-DNN paradigms, to improve the accuracy of models controlling normalization and frequency of communications, and to boost data privacy though generative adversarial networks. Prerequisites: Intermediate understanding of machine learning methods and distributed & parallel computing.},
howpublished = {NVidia GTC 2023},
keywords = {across, admire, epi, eumaster4hpc, eupex, eupilot, hpc4ai, space, textarossa},
pubstate = {published},
tppubtype = {misc}
}
Decentralized machine learning (DML) enables collaborative machine learning without centralized input data. Federated learning (FL) and edge inference (EI) are examples of DML. Collaboration naturally happens at the edge of a distributed system with inherently distributed data. While tools for DML are starting to flourish, much needs to be done to get more flexible and portable tools to experiment with novel techniques, non-fully connected topologies, multiple data domains, and asynchronous collaboration schemes. We'll present recent advances in DML, aiming to improve usability in data centers and, at the edge, to widen the class of models extending FL to non-DNN paradigms, to improve the accuracy of models controlling normalization and frequency of communications, and to boost data privacy though generative adversarial networks. Prerequisites: Intermediate understanding of machine learning methods and distributed & parallel computing.
Marco Aldinucci
HPC4AI: The Research on AI beyond the public cloud Miscellaneous
CENTAI kick-off meeting, 2023.
Links | BibTeX | Tags: across, admire, brainteaser, epi, eumaster4hpc, eupex, eupilot, hpc4ai, space, textarossa
@misc{23:CENTAI:hpc4ai,
title = {HPC4AI: The Research on AI beyond the public cloud},
author = {Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/PZXjPm8sfKTmTGb},
year = {2023},
date = {2023-03-01},
address = {Torino, Italy},
howpublished = {CENTAI kick-off meeting},
keywords = {across, admire, brainteaser, epi, eumaster4hpc, eupex, eupilot, hpc4ai, space, textarossa},
pubstate = {published},
tppubtype = {misc}
}
Marco Aldinucci
From HPC4AI to ICSC living lab: Where systems are the research Miscellaneous
Dell Advanced Computing Workshop 2023: HPC and Beyond, 2023.
Links | BibTeX | Tags: admire, eupex, eupilot, hpc4ai, textarossa
@misc{23:Dell:hpc4ai,
title = {From HPC4AI to ICSC living lab: Where systems are the research},
author = {Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/M5QRJyDxyxokcfL},
year = {2023},
date = {2023-02-01},
address = {Bologna, Italy},
howpublished = {Dell Advanced Computing Workshop 2023: HPC and Beyond},
keywords = {admire, eupex, eupilot, hpc4ai, textarossa},
pubstate = {published},
tppubtype = {misc}
}
Gianluca Mittone
Paving the way to innovative tools for Federated Learning Miscellaneous
2023 HiPEAC Conference, 2023, (Invited talk).
Abstract | Links | BibTeX | Tags: eupilot
@misc{23:hipeac,
title = {Paving the way to innovative tools for Federated Learning},
author = {Gianluca Mittone},
url = {https://datacloud.di.unito.it/index.php/s/2GtxPidHq79RTzA},
year = {2023},
date = {2023-02-01},
address = {Toulouse, France},
abstract = {Since its debut in 2016, Federated Learning (FL) has been tied to the inner workings of Deep Neural Networks (DNNs). On the one hand, this allowed its development and widespread use as DNNs proliferated. On the other hand, it neglected all those scenarios in which using DNNs is not possible or advantageous. The fact that most current FL frameworks only allow training DNNs reinforces this problem. To address the lack of FL solutions for non-DNN-based use cases, we propose MAFL (Model-Agnostic Federated Learning). MAFL marries a model-agnostic FL algorithm, AdaBoost.F, with an open industry-grade FL framework: Intel® OpenFL. MAFL is the first FL system not tied to any specific type of machine learning model, allowing exploration of FL scenarios beyond DNNs and trees. Furthermore, tools for DML (especially FL) are starting to flourish, many are not flexible and portable enough to experiment with novel systems (e.g., RISC-V), non-fully connected topologies, and asynchronous collaboration schemes. We overcome these limitations via a domain-specific language allowing to map DML schemes to an underlying middleware, i.e. the ff parallel programming library. As a byproduct, we introduce a RISC-V porting of the PyTorch framework, the first publicly available to our knowledge.},
howpublished = {2023 HiPEAC Conference},
note = {Invited talk},
keywords = {eupilot},
pubstate = {published},
tppubtype = {misc}
}
Since its debut in 2016, Federated Learning (FL) has been tied to the inner workings of Deep Neural Networks (DNNs). On the one hand, this allowed its development and widespread use as DNNs proliferated. On the other hand, it neglected all those scenarios in which using DNNs is not possible or advantageous. The fact that most current FL frameworks only allow training DNNs reinforces this problem. To address the lack of FL solutions for non-DNN-based use cases, we propose MAFL (Model-Agnostic Federated Learning). MAFL marries a model-agnostic FL algorithm, AdaBoost.F, with an open industry-grade FL framework: Intel® OpenFL. MAFL is the first FL system not tied to any specific type of machine learning model, allowing exploration of FL scenarios beyond DNNs and trees. Furthermore, tools for DML (especially FL) are starting to flourish, many are not flexible and portable enough to experiment with novel systems (e.g., RISC-V), non-fully connected topologies, and asynchronous collaboration schemes. We overcome these limitations via a domain-specific language allowing to map DML schemes to an underlying middleware, i.e. the ff parallel programming library. As a byproduct, we introduce a RISC-V porting of the PyTorch framework, the first publicly available to our knowledge.
2022
Bruno Casella
Benchmarking FedAvg and FedCurv for Image Classification Tasks Miscellaneous
ITADATA, 2022.
Abstract | Links | BibTeX | Tags: eupilot
@misc{22:itadata,
title = {Benchmarking FedAvg and FedCurv for Image Classification Tasks},
author = {Bruno Casella},
url = {https://datacloud.di.unito.it/index.php/s/6XaEXnAowRrAHGL},
year = {2022},
date = {2022-09-01},
address = {Milan, Italy},
abstract = {Presentation of the paper "Benchmarking FedAvg and FedCurv for Image Classification Tasks" to the first italian conference on Big Data and Data Science},
howpublished = {ITADATA},
keywords = {eupilot},
pubstate = {published},
tppubtype = {misc}
}
Presentation of the paper "Benchmarking FedAvg and FedCurv for Image Classification Tasks" to the first italian conference on Big Data and Data Science
Marco Aldinucci
Il calcolo parallelo: una storia di metodi e algoritmi raccontata dalle macchine Miscellaneous
Olimpiadi di Informatica, 2022, (Invited talk).
Abstract | Links | BibTeX | Tags: across, admire, eumaster4hpc, eupex, eupilot, textarossa
@misc{22:olimpiadi:cs,
title = {Il calcolo parallelo: una storia di metodi e algoritmi raccontata dalle macchine},
author = {Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/7ZdfLkn3NetzXCN},
year = {2022},
date = {2022-09-01},
address = {Biella, Italy},
abstract = {Lectio Magistralis alle finali nazionali delle Olimpiadi di Informatica 2022},
howpublished = {Olimpiadi di Informatica},
note = {Invited talk},
keywords = {across, admire, eumaster4hpc, eupex, eupilot, textarossa},
pubstate = {published},
tppubtype = {misc}
}
Lectio Magistralis alle finali nazionali delle Olimpiadi di Informatica 2022
Marco Aldinucci
La convergenza HPC-cloud è l'anello mancante tra il calcolo scientifico e l'IA applicata Miscellaneous
Intelligenza Artificiale e Business Applications, 2022, (Invited talk).
Abstract | Links | BibTeX | Tags: across, admire, eumaster4hpc, eupex, eupilot, textarossa
@misc{22:soiel:ai,
title = {La convergenza HPC-cloud è l'anello mancante tra il calcolo scientifico e l'IA applicata},
author = {Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/xCQSqJ8bCKCXMK9},
year = {2022},
date = {2022-09-01},
address = {Virtual event},
abstract = {Innanzitutto, le infrastrutture HPC stanno adottando le GPU per il loro rapporto prestazioni per watt superiore rispetto ai multicore generici. In secondo luogo, i flussi di lavoro scientifici di prossima generazione stanno integrando passaggi basati sull'intelligenza artificiale per la loro precisione nell'approssimazione e nell'analisi di fenomeni complessi. In terzo luogo, l'IA e in particolare il Machine Learning (ML) rappresentano un carico di lavoro perfetto per le GPU in termini di prestazioni e tempo di sviluppo. Oggi non possiamo ancora chiudere il cerchio eseguendo senza problemi carichi di lavoro scientifici abilitati all'intelligenza artificiale nelle infrastrutture HPC perché il loro software di sistema e gli strumenti di sviluppo non sono progettati per i carichi di lavoro moderni, come i framework ML progettati per il cloud. È probabile che la convergenza HPC-cloud colmi il divario. Nel talk verranno presentate le infrastrutture e gli strumenti sviluppati all'Università di Torino per la convergenza HPC-cloud (es. HPC4AI, StreamFlow, CAPIO, Jupyter-workflow) e come sono stati utilizzati per le applicazioni di intelligenza artificiale, come la diagnosi spiegabile di polmonite COVID-19 e la tutela della privacy AI. L'esperienza maturata nella progettazione e gestione di HPC4AI costituisce il cuore della progettazione del laboratorio di contaminazione del "FutureHPC" di Torino secondo il Centro Nazionale "HPC, BigData e Quantum Computing" finanziato dal PNRR con 320M€ che dovrebbe essere operativo dal 1 settembre 2022. L'obiettivo finale del laboratorio di contaminazione è sviluppare relazioni e collaborazioni tra industria e università.},
howpublished = {Intelligenza Artificiale e Business Applications},
note = {Invited talk},
keywords = {across, admire, eumaster4hpc, eupex, eupilot, textarossa},
pubstate = {published},
tppubtype = {misc}
}
Innanzitutto, le infrastrutture HPC stanno adottando le GPU per il loro rapporto prestazioni per watt superiore rispetto ai multicore generici. In secondo luogo, i flussi di lavoro scientifici di prossima generazione stanno integrando passaggi basati sull'intelligenza artificiale per la loro precisione nell'approssimazione e nell'analisi di fenomeni complessi. In terzo luogo, l'IA e in particolare il Machine Learning (ML) rappresentano un carico di lavoro perfetto per le GPU in termini di prestazioni e tempo di sviluppo. Oggi non possiamo ancora chiudere il cerchio eseguendo senza problemi carichi di lavoro scientifici abilitati all'intelligenza artificiale nelle infrastrutture HPC perché il loro software di sistema e gli strumenti di sviluppo non sono progettati per i carichi di lavoro moderni, come i framework ML progettati per il cloud. È probabile che la convergenza HPC-cloud colmi il divario. Nel talk verranno presentate le infrastrutture e gli strumenti sviluppati all'Università di Torino per la convergenza HPC-cloud (es. HPC4AI, StreamFlow, CAPIO, Jupyter-workflow) e come sono stati utilizzati per le applicazioni di intelligenza artificiale, come la diagnosi spiegabile di polmonite COVID-19 e la tutela della privacy AI. L'esperienza maturata nella progettazione e gestione di HPC4AI costituisce il cuore della progettazione del laboratorio di contaminazione del "FutureHPC" di Torino secondo il Centro Nazionale "HPC, BigData e Quantum Computing" finanziato dal PNRR con 320M€ che dovrebbe essere operativo dal 1 settembre 2022. L'obiettivo finale del laboratorio di contaminazione è sviluppare relazioni e collaborazioni tra industria e università.
Iacopo Colonnelli, Barbara Cantalupo, Doriana Medić, Marco Aldinucci
Hybrid workflows for heterogeneous distributed computing Miscellaneous
3rd Italian Workshop on HPC (ITWSHPC), 2022.
Links | BibTeX | Tags: across, admire, eumaster4hpc, eupex, eupilot, textarossa
@misc{22:itwshpc,
title = {Hybrid workflows for heterogeneous distributed computing},
author = {Iacopo Colonnelli and Barbara Cantalupo and Doriana Medić and Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/ienbcA2DJ26aioE},
year = {2022},
date = {2022-09-01},
address = {Torino, Italy},
howpublished = {3rd Italian Workshop on HPC (ITWSHPC)},
keywords = {across, admire, eumaster4hpc, eupex, eupilot, textarossa},
pubstate = {published},
tppubtype = {misc}
}
Iacopo Colonnelli, Marco Aldinucci
CINI HPC-KTT: HPC Key Technologies and Tools National Lab Miscellaneous
NVIDIA HPC Roundtable, 2022, (Invited talk).
Links | BibTeX | Tags: across, admire, eumaster4hpc, eupex, eupilot, textarossa
@misc{22:nvidia_hpc_roundtable,
title = {CINI HPC-KTT: HPC Key Technologies and Tools National Lab},
author = {Iacopo Colonnelli and Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/9EQniZ2dGzdJ26f},
year = {2022},
date = {2022-09-01},
address = {Casalecchio di Reno, Italy},
howpublished = {NVIDIA HPC Roundtable},
note = {Invited talk},
keywords = {across, admire, eumaster4hpc, eupex, eupilot, textarossa},
pubstate = {published},
tppubtype = {misc}
}
Marco Aldinucci
EuroHPC and the Italian HPC ecosystem Miscellaneous
Critical Infrastructure Protection Forum - EuroCC Romania, 2022, (Invited talk).
Abstract | Links | BibTeX | Tags: across, admire, eumaster4hpc, eupex, eupilot, icsc, textarossa
@misc{22:cip:romania,
title = {EuroHPC and the Italian HPC ecosystem},
author = {Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/5dFFoNsZzwTzQkn},
year = {2022},
date = {2022-06-01},
address = {Bucharest, Romania},
abstract = {The talk presents the main investments currently ongoing in Italy in the HPC area as well as the activity of Italian stakeholders within EuroHPC. The novel Italian National Centre on HPC (ICSC) is introduced.},
howpublished = {Critical Infrastructure Protection Forum - EuroCC Romania},
note = {Invited talk},
keywords = {across, admire, eumaster4hpc, eupex, eupilot, icsc, textarossa},
pubstate = {published},
tppubtype = {misc}
}
The talk presents the main investments currently ongoing in Italy in the HPC area as well as the activity of Italian stakeholders within EuroHPC. The novel Italian National Centre on HPC (ICSC) is introduced.
Marco Aldinucci
The Italian HPC ecosystem and the next generation of EuroHPC CoE Miscellaneous
EuroHPC EoCoE final summit, 2022, (Invited talk).
Abstract | Links | BibTeX | Tags: across, admire, eumaster4hpc, eupex, eupilot, icsc, textarossa
@misc{22:eocoe:summit,
title = {The Italian HPC ecosystem and the next generation of EuroHPC CoE},
author = {Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/AH5Ms3NekeoEooB},
year = {2022},
date = {2022-06-01},
address = {Napoli, Italy},
abstract = {The talk presents the main investments currently ongoing in Italy in the HPC area as well as the activity of Italian stakeholders within EuroHPC. The novel Italian National Centre on HPC (ICSC) is introduced.},
howpublished = {EuroHPC EoCoE final summit},
note = {Invited talk},
keywords = {across, admire, eumaster4hpc, eupex, eupilot, icsc, textarossa},
pubstate = {published},
tppubtype = {misc}
}
The talk presents the main investments currently ongoing in Italy in the HPC area as well as the activity of Italian stakeholders within EuroHPC. The novel Italian National Centre on HPC (ICSC) is introduced.
Marco Aldinucci
HPC-cloud convergence is the missing link between scientific computing and applied-AI Miscellaneous
Machine Learning for Astrophysics (ML4ASTRO), 2022, (Keynote talk).
Abstract | Links | BibTeX | Tags: across, deephealth, eupex, eupilot
@misc{22:ml4astrotalk,
title = {HPC-cloud convergence is the missing link between scientific computing and applied-AI},
author = {Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/2SGswkcip7MoMoH},
year = {2022},
date = {2022-06-01},
address = {Catania, Italy},
abstract = {First, HPC infrastructures are embracing GPUs for their superior performance-per-watt ratio against general-purpose multicores. Second, the next-generation scientific workflows are integrating AI-based steps for their accuracy in approximating and analyzing complex phenomena. Third, AI and specifically Machine Learning (ML), is a perfect workload for GPUs in terms of performance and development time. Today, we cannot still close the circle seamlessly running AI-enabled scientific workloads into HPC infrastructures because their system software and development tools are not designed for modern workloads, such as ML frameworks designed for the cloud. HPC-cloud convergence is likely to bridge the gap. In the talk, we will present Streamflow and CAPIO, two development tools for HPC-cloud convergence.},
howpublished = {Machine Learning for Astrophysics (ML4ASTRO)},
note = {Keynote talk},
keywords = {across, deephealth, eupex, eupilot},
pubstate = {published},
tppubtype = {misc}
}
First, HPC infrastructures are embracing GPUs for their superior performance-per-watt ratio against general-purpose multicores. Second, the next-generation scientific workflows are integrating AI-based steps for their accuracy in approximating and analyzing complex phenomena. Third, AI and specifically Machine Learning (ML), is a perfect workload for GPUs in terms of performance and development time. Today, we cannot still close the circle seamlessly running AI-enabled scientific workloads into HPC infrastructures because their system software and development tools are not designed for modern workloads, such as ML frameworks designed for the cloud. HPC-cloud convergence is likely to bridge the gap. In the talk, we will present Streamflow and CAPIO, two development tools for HPC-cloud convergence.
Marco Aldinucci
Cognitive continuum: a game theoretical approach Miscellaneous
HiPEAC Vision meeting, Brussels, 16 May 2022, 2022.
Abstract | Links | BibTeX | Tags: across, admire, brainteaser, eumaster4hpc, eupex, eupilot, textarossa
@misc{22:hipeacvision:fl,
title = {Cognitive continuum: a game theoretical approach},
author = {Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/453HWfmrQyo7j9E},
year = {2022},
date = {2022-05-01},
address = {Brussels, Belgium},
abstract = {Cognitive continuum: a game theoretical approach, (maybe) data operations are too basic: read, write, copy, remove … The talk is aimed to contribute to the forthcoming HiPEAC Vision document},
howpublished = {HiPEAC Vision meeting, Brussels, 16 May 2022},
keywords = {across, admire, brainteaser, eumaster4hpc, eupex, eupilot, textarossa},
pubstate = {published},
tppubtype = {misc}
}
Cognitive continuum: a game theoretical approach, (maybe) data operations are too basic: read, write, copy, remove … The talk is aimed to contribute to the forthcoming HiPEAC Vision document
2021
Marco Aldinucci
The Italian research on HPC key technologies across EuroHPC Miscellaneous
2021.
Abstract | Links | BibTeX | Tags: across, admire, eupex, eupilot, textarossa
@misc{21:CINI_acm_CF_talk,
title = {The Italian research on HPC key technologies across EuroHPC},
author = {Marco Aldinucci},
url = {https://datacloud.di.unito.it/index.php/s/3ZYmDbEm84rbB9k},
year = {2021},
date = {2021-05-01},
booktitle = {ACM Computing Frontiers},
publisher = {ACM},
address = {Virtual Conference, Italy},
abstract = {High-Performance Computing (HPC) is one of the strategic priorities for research and innovation worldwide due to its relevance for industrial and scientific applications. We envision HPC as composed of three pillars: infrastructures, applications, and key technologies and tools. While infrastructures are by construction centralized in large-scale HPC centers, and applications are generally within the purview of domain-specific organizations, key technologies fall in an intermediate case where coordination is needed, but design and development are often decentralized. A large group of Italian researchers has started a dedicated laboratory within the National Interuniversity Consortium for Informatics (CINI) to address this challenge. The laboratory, albeit young, has managed to succeed in its first attempts to propose a coordinated approach to HPC research within the EuroHPC Joint Undertaking, participating in the calls 2019-20 to five successful proposals for an aggregate total cost of 95M Euro. In this paper, we outline the working group's scope and goals and provide an overview of the five funded projects, which become fully operational in March 2021, and cover a selection of key technologies provided by the working group partners, highlighting their usage development within the projects.},
keywords = {across, admire, eupex, eupilot, textarossa},
pubstate = {published},
tppubtype = {misc}
}
High-Performance Computing (HPC) is one of the strategic priorities for research and innovation worldwide due to its relevance for industrial and scientific applications. We envision HPC as composed of three pillars: infrastructures, applications, and key technologies and tools. While infrastructures are by construction centralized in large-scale HPC centers, and applications are generally within the purview of domain-specific organizations, key technologies fall in an intermediate case where coordination is needed, but design and development are often decentralized. A large group of Italian researchers has started a dedicated laboratory within the National Interuniversity Consortium for Informatics (CINI) to address this challenge. The laboratory, albeit young, has managed to succeed in its first attempts to propose a coordinated approach to HPC research within the EuroHPC Joint Undertaking, participating in the calls 2019-20 to five successful proposals for an aggregate total cost of 95M Euro. In this paper, we outline the working group's scope and goals and provide an overview of the five funded projects, which become fully operational in March 2021, and cover a selection of key technologies provided by the working group partners, highlighting their usage development within the projects.