Zobrazeno 1 - 10
of 24 869
pro vyhledávání: '"Keyes A"'
Publikováno v:
Neural Information Processing Systems (NeurIPS). Machine Learning with New Compute Paradigms (MLNCP) Workshop, October 2024
We present a novel approach for accelerating AI performance by leveraging Anderson extrapolation, a vector-to-vector mapping technique based on a window of historical iterations. By identifying the crossover point where a mixing penalty is incurred,
Externí odkaz:
http://arxiv.org/abs/2410.19460
This paper explores the performance optimization of out-of-core (OOC) Cholesky factorization on shared-memory systems equipped with multiple GPUs. We employ fine-grained computational tasks to expose concurrency while creating opportunities to overla
Externí odkaz:
http://arxiv.org/abs/2410.09819
We develop a theory for the electrical and thermal transverse linear response functions such as the Hall, Nernst and thermal Hall effects in magnetic materials that harbor topological spin textures like skyrmions. In addition to the ordinary transver
Externí odkaz:
http://arxiv.org/abs/2409.04376
Autor:
Ltaief, Hatem, Alomairy, Rabab, Cao, Qinglei, Ren, Jie, Slim, Lotfi, Kurth, Thorsten, Dorschner, Benedikt, Bougouffa, Salim, Abdelkhalak, Rached, Keyes, David E.
We exploit the widening margin in tensor-core performance between [FP64/FP32/FP16/INT8,FP64/FP32/FP16/FP8/INT8] on NVIDIA [Ampere,Hopper] GPUs to boost the performance of output accuracy-preserving mixed-precision computation of Genome-Wide Associati
Externí odkaz:
http://arxiv.org/abs/2409.01712
Autor:
Abdulah, Sameh, Baker, Allison H., Bosilca, George, Cao, Qinglei, Castruccio, Stefano, Genton, Marc G., Keyes, David E., Khalid, Zubair, Ltaief, Hatem, Song, Yan, Stenchikov, Georgiy L., Sun, Ying
We present the design and scalable implementation of an exascale climate emulator for addressing the escalating computational and storage requirements of high-resolution Earth System Model simulations. We utilize the spherical harmonic transform to s
Externí odkaz:
http://arxiv.org/abs/2408.04440
Publikováno v:
International Conference on Machine Learning (ICML). Machine Learning for Life and Material Science (ML4LMS) Workshop, May 2024
Deep AndersoNN accelerates AI by exploiting the continuum limit as the number of explicit layers in a neural network approaches infinity and can be taken as a single implicit layer, known as a deep equilibrium model. Solving for deep equilibrium mode
Externí odkaz:
http://arxiv.org/abs/2407.19724
Autor:
Craig, Erin, Keyes, Timothy, Sarno, Jolanda, Zaslavsky, Maxim, Nolan, Garry, Davis, Kara, Hastie, Trevor, Tibshirani, Robert
Single-cell datasets often lack individual cell labels, making it challenging to identify cells associated with disease. To address this, we introduce Mixture Modeling for Multiple Instance Learning (MMIL), an expectation maximization method that ena
Externí odkaz:
http://arxiv.org/abs/2406.08322
Personal names simultaneously differentiate individuals and categorize them in ways that are important in a given society. While the natural language processing community has thus associated personal names with sociodemographic characteristics in a v
Externí odkaz:
http://arxiv.org/abs/2405.17159
Autor:
Zhang, Xiran, Abdulah, Sameh, Cao, Jian, Ltaief, Hatem, Sun, Ying, Genton, Marc G., Keyes, David E.
Addressing the statistical challenge of computing the multivariate normal (MVN) probability in high dimensions holds significant potential for enhancing various applications. One common way to compute high-dimensional MVN probabilities is the Separat
Externí odkaz:
http://arxiv.org/abs/2405.14892
Autor:
Beneish, Lea, Keyes, Christopher
A cubic hypersurface in $\mathbb{P}^n$ defined over $\mathbb{Q}$ is given by the vanishing locus of a cubic form $f$ in $n+1$ variables. It is conjectured that when $n \geq 4$, such cubic hypersurfaces satisfy the Hasse principle. This is now known t
Externí odkaz:
http://arxiv.org/abs/2405.06584