FPGA MPS Tensor Network Contractions - ON-1085

Genre de projet: Recherche
Discipline(s) souhaitée(s): Génie - informatique / électrique, Génie, Informatique, Sciences mathématiques
Entreprise: Multiverse Computing
Durée du projet: Plus d’un an
Date souhaitée de début: Dès que possible
Langue exigée: Anglais
Emplacement(s): Toronto, ON, Canada
Nombre de postes: 1
Niveau de scolarité désiré: MaîtriseDoctoratRecherche postdoctorale
Ouvert aux candidatures de personnes inscrites à un établissement à l’extérieur du Canada: No

Au sujet de l’entreprise: 

Multiverse is a well funded deep-tech Canadian company. We are one of the few world companies working with Quantum Computing. We provide hyper-efficient software for companies from the financial industry wanting to gain an edge with quantum computing and artificial intelligence. Our main verticals are fraud detection, credit scoring assessment, and financial optimization.

Our team of experts is world-renowned for innovative approaches to intractable financial and macro-economics problems. We work with quantum hardware and quantum inspired methods to build machine learning solutions which exceed the predictive power of the current best solutions.

We are applying to Mitacs to allow us to expand our R&D team in Canada to tackle problems of huge commercial impact and expand our expertise.

Veuillez décrire le projet.: 

The objective of this project is to leverage Tensor Neural Networks and Quantum-inspired methodologies to enhance the development and efficiency of machine learning models. A key focus is on optimizing the compression of large-scale models, including Large Language Models (LLMs), Computer Vision models, and Text-to-Speech models, to improve computational efficiency while maintaining high performance. By integrating advanced techniques from quantum computing and tensor-based approaches, the project aims to enable more efficient model training, storage, and deployment, making AI systems more scalable and accessible.
● Understand and run current FPGA hardware for tensor contraction
● Design data flow for MPS tensor contractions to maximize parallelism
● Apply tensor contraction modules for MPS data flow architecture
● Interface with classical program from Python.

Expertise ou compétences exigées: 

The ideal candidate should have expertise in deep learning, particularly with Tensor Neural Networks (TNN) and experience in tensor networks for machine learning applications. Strong skills in programming proficiency in Python for developing machine learning frameworks and software interfaces. Knowledge of numerical methods and optimization algorithms for tensor network contraction is required. The candidate should also be adept at benchmarking and performance evaluation of computational methods. Strong scientific documentation skills are necessary. Optional assets include familiarity with quantum computing (e.g., Qiskit) and experience in highperformance computing (HPC).