PIRSA:22110028

Video URL

https://pirsa.org/22110028

Computability in QFT and tropical QFT

Borinsky, M. (2022). Computability in QFT and tropical QFT. Perimeter Institute for Theoretical Physics. https://pirsa.org/22110028

Borinsky, Michael. Computability in QFT and tropical QFT. Perimeter Institute for Theoretical Physics, Nov. 22, 2022, https://pirsa.org/22110028 

          @misc{ scivideos_PIRSA:22110028,
            doi = {10.48660/22110028},
            url = {https://pirsa.org/22110028},
            author = {Borinsky, Michael},
            keywords = {Quantum Fields and Strings},
            language = {en},
            title = {Computability in QFT and tropical QFT},
            publisher = {Perimeter Institute for Theoretical Physics},
            year = {2022},
            month = {nov},
            note = {PIRSA:22110028 see, \url{https://scivideos.org/index.php/pirsa/22110028}}
          }

Michael Borinsky ETH Zurich

Talk numberPIRSA:22110028
DOI10.48660/22110028
Source RepositoryPIRSA
Collection
Talk Type Scientific Series
Subject

Abstract

In general, observables in QFTs can only be computed perturbatively using Feynman integrals. In this talk, which is based on arXiv:2008.12310 and arXiv:2204.06414, I will address some questions on the computability of such observables. By looking at QFT through the lens of tropical geometry, one can see that the runtime of such computations is dominated by the time it takes to understand the structure of certain polytopes. In the case of scalar QFTs on Euclidean space the relevant polytopes turn out to be generalized permutahedra, whose structure is well-understood thanks to the works of Postnikov, Aguiar and Ardila.  Using these insights, results in a new algorithm for Feynman integral evaluation that exceeds the capabilities of existing methods by multiple orders of magnitude, while being easy to implement.  I will also briefly discuss current wip that extends these findings to QFTs on Minkowski spacetime.

Zoom link:  https://pitp.zoom.us/j/93629580865?pwd=WnJ6aUVpckFZWGVDL0NqMTFrWlhBQT09