SubsidieMeestersAnalyzing and Exploiting Inexactness in Exascale Matrix Computations
This project aims to develop a holistic framework for analyzing and exploiting multiple sources of inexactness in matrix computations to enhance algorithm performance and accuracy for exascale applications.
Projectdetails
Introduction
Scientific computing inherently involves multiple sources of inexactness, from discretization or simplification of the problem, to noisy data, to finite precision rounding errors, to approximations made to increase parallelism, to stopping computations intentionally to improve efficiency. The standard state-of-the-art approach is to analyze different sources of error separately.
Problem Statement
There is currently no solid foundation or systematic approach for combining multiple sources of inexactness together and studying their interaction. Developing reliable approaches for exascale requires filling this gap, which must start with establishing a new rigorous foundation for analyzing multiple sources of error in matrix computations. Without this basis, the quest for efficiency in areas vitally depending on matrix computations, including, for example, data science and machine learning, will remain reliant on a trial-and-error approach.
Project Objectives
This project aims to break the current modular approach to the analysis and design of algorithms for matrix computations by understanding how different sources of inexactness interact while being propagated through a computation and their effect on numerical behavior and solution quality. Our holistic approach, rooted in rigorous theoretical analysis, will reveal opportunities for developing new algorithms for exascale problems that exploit inexactness to balance performance and accuracy.
The project is structured around four fundamental objectives:
- WP1: Analysis of exascale matrix computations subject to multiple sources of inexactness
- WP2: Development of new algorithms that exploit inexactness that are both fast and provably accurate
- WP3: Making error analysis of exascale computations meaningful in practice
- WP4: Exploring emerging sources of inexactness beyond the exascale era
Conclusion
Our approach will lead to new methodologies that can change current paradigms.
Financiële details & Tijdlijn
Financiële details
| Subsidiebedrag | € 1.496.085 |
| Totale projectbegroting | € 1.496.085 |
Tijdlijn
| Startdatum | 1-3-2023 |
| Einddatum | 29-2-2028 |
| Subsidiejaar | 2023 |
Partners & Locaties
Projectpartners
- UNIVERZITA KARLOVApenvoerder
Land(en)
Vergelijkbare projecten binnen European Research Council
| Project | Regeling | Bedrag | Jaar | Actie |
|---|---|---|---|---|
Systematic and computer-aided performance certification for numerical optimizationThe project aims to enhance theoretical foundations of numerical optimization to bridge the gap between theory and practice, developing robust algorithms and certification tools for complex applications. | ERC Starting... | € 1.497.650 | 2024 | Details |
Exact and Approximate Computation of Tensors and PolynomialsThis project aims to tackle fundamental challenges in polynomial computation and manipulation, seeking breakthroughs in complexity, algorithms, and quantum information theory. | ERC Advanced... | € 2.335.000 | 2024 | Details |
Error-correcting Codes and ComputationThe project aims to design advanced error-correcting codes that optimize redundancy and error-resilience while enabling fast algorithms, with applications in computational efficiency and cryptography. | ERC Starting... | € 1.489.375 | 2023 | Details |
Solving differential equations fast, precisely, and reliablyThis project aims to enhance the speed and reliability of solving differential equations by developing new computational methods and open-source libraries for both numeric and symbolic solutions. | ERC Advanced... | € 2.396.711 | 2024 | Details |
Predictive algorithms for simulating quantum materialsThis project aims to develop advanced predictive algorithms for quantum many-body systems by integrating field-theory methods with tensor techniques and machine learning to enhance understanding of quantum materials. | ERC Advanced... | € 3.499.299 | 2025 | Details |
Systematic and computer-aided performance certification for numerical optimization
The project aims to enhance theoretical foundations of numerical optimization to bridge the gap between theory and practice, developing robust algorithms and certification tools for complex applications.
Exact and Approximate Computation of Tensors and Polynomials
This project aims to tackle fundamental challenges in polynomial computation and manipulation, seeking breakthroughs in complexity, algorithms, and quantum information theory.
Error-correcting Codes and Computation
The project aims to design advanced error-correcting codes that optimize redundancy and error-resilience while enabling fast algorithms, with applications in computational efficiency and cryptography.
Solving differential equations fast, precisely, and reliably
This project aims to enhance the speed and reliability of solving differential equations by developing new computational methods and open-source libraries for both numeric and symbolic solutions.
Predictive algorithms for simulating quantum materials
This project aims to develop advanced predictive algorithms for quantum many-body systems by integrating field-theory methods with tensor techniques and machine learning to enhance understanding of quantum materials.