4 papers
Why Smaller Is Slower? Dimensional Misalignment in Compressed LLMs
Jihao Xin, Tian Lyu, Qilong Pan +2
Post-training compression reduces LLM parameter counts but often produces irregular tensor dimensions that degrade GPU performance -- a phenomenon we call \emph{dimensional misalig…
Scaled Block Vecchia Approximation for High-Dimensional Gaussian Process Emulation on GPUs
Qilong Pan, Sameh Abdulah, Mustafa Abduljabbar +8
Emulating computationally intensive scientific simulations is crucial for enabling uncertainty quantification, optimization, and informed decision-making at scale. Gaussian Process…
Block Vecchia Approximation for Scalable and Efficient Gaussian Process Computations
Qilong Pan, Sameh Abdulah, Marc G. Genton +1
Gaussian Processes (GPs) are vital for modeling and predicting irregularly-spaced, large geospatial datasets. However, their computations often pose significant challenges in large…
GPU-Accelerated Vecchia Approximations of Gaussian Processes for Geospatial Data using Batched Matrix Computations
Qilong Pan, Sameh Abdulah, Marc G. Genton +3
Gaussian processes (GPs) are commonly used for geospatial analysis, but they suffer from high computational complexity when dealing with massive data. For instance, the log-likelih…