GPU Accelerated Sampling and Model Comparison

Add to your list(s) Download to your calendar using vCal

• David Yallup / IoA
• Wednesday 04 June 2025, 13:40-14:05
• The Hoyle Lecture Theatre + Zoom .

If you have a question about this talk, please contact .

This talk introduces a natively vectorized implementation of the Nested Sampling algorithm, enabling deployment of the entire inference process onto GPUs for massive acceleration. I will start by reviewing the benefits, and necessity, of the paradigm shift towards vectorized compute in the physical sciences. After a brief review of the how (and why) of Bayesian inference in Astronomy and Cosmology, I will then explore the nuances and challenges of taking some of the widely used inference algorithms within this community, in particular nested sampling, to the GPU accelerated frontier. Lastly I’ll present some practical benefit that this speedup can bring and comment on how this technical development can help push the boundaries of what we can achieve in the physical sciences.

This talk is part of the Institute of Astronomy Seminars series.

This talk is included in these lists:

• Cambridge Astronomy Talks
• Combined External Astrophysics Talks DAMTP
• Cosmology lists
• Cosmology, Astrophysics and General Relativity
• Institute of Astronomy Seminars
• Institute of Astronomy Talk Lists
• Kavli Institute for Cosmology Talk Lists
• Priscilla
• The Hoyle Lecture Theatre + Zoom

Note that ex-directory lists are not shown.