How do you simulate chiral gauge theories on a computer?

Question

David Tong and Lubos Motl have argued that our universe can't possibly be a digital computer simulation because chiral gauge theories can't be discretized, and the Standard Model is a chiral gauge theory. Certainly, you can't regulate them on a lattice. However, that doesn't mean they're not limit computable. There are only two alternatives. Either chiral gauge theories are uncomputable (extremely unlikely), or they can be simulated on a digital computer. How do you simulate a chiral gauge theory on a digital computer? Attempts by Erich Poppitz have fallen a bit short of the goal.

This post imported from StackExchange Physics at 2014-04-05 03:00 (UCT), posted by SE-user Tefaidr

Comments

Define "digital computer"... as it stands, this question is subjective. The argument for "our universe can't be a digital computer" on face-value is simply the statement "a definite integral can only be approximated by a discrete finite sum".

This post imported from StackExchange Physics at 2014-04-05 03:00 (UCT), posted by SE-user Chris Gerig

---

If you're asking what's better than domain wall fermions, then you're asking an open question.

This post imported from StackExchange Physics at 2014-04-05 03:00 (UCT), posted by SE-user user1504

---

Is there something involved, which has a decission on what to simulate? Are there different possible things to be simulated?

This post imported from StackExchange Physics at 2014-04-05 03:00 (UCT), posted by SE-user NiftyKitty95

---

Just for clarification, is the problem with discretization that you're referring to the one described in sections 1 and 2 here?

This post imported from StackExchange Physics at 2014-04-05 03:00 (UCT), posted by SE-user twistor59

---

@ChrisGerig I think we can safely suggest that in the term digital computer, Turing Machine is implied. However, I am not suggesting this is the only problem with this question...

This post imported from StackExchange Physics at 2014-04-05 03:00 (UCT), posted by SE-user Killercam

Answer 1

Overlap fermion approach may be the answer. Ounce a theory is defined on a lattice, it can be simulated by a computer that we already have. Here is a review on overlap fermion approach:

Tata lectures on overlap fermions arXiv:1103.4588

R. Narayanan

Overlap formalism deals with the construction of chiral gauge theories on the lattice. These set of lectures provide a pedagogical introduction to the subject with emphasis on chiral anomalies and gauge field topology. Subtleties associated with the generating functional for gauge theories coupled to chiral fermions are discussed.

==== A new result ===

One can simulate any anomaly-free chiral gauge theories on a computer by simply put it on lattice and turn on a proper interaction. See my new papers http://arxiv.org/abs/1305.1045 and http://arxiv.org/abs/1303.1803

The paper http://arxiv.org/abs/1305.1045 was rejected by PRL (see the referee's comments and my reply http://bbs.sciencenet.cn/home.php?mod=space&uid=1116346&do=blog&id=736247 ). It is now published in CPL.

This post imported from StackExchange Physics at 2014-04-05 03:00 (UCT), posted by SE-user Xiao-Gang Wen

Answer 2

See also this paper: arxiv-1307.7480: Lattice Non-Perturbative Definition of 1+1D Anomaly-Free Chiral Fermions and Bosons. This paper follows Prof. Wen's general thinking and provide a proof between the following two conditions:

"Topological Boundary (Gapping) Conditions"

is equivalent to

"t' Hooft anomaly matching conditions"

The proof is given for the case of the a theory with a U(1) symmetry and in 1+1D.

Using this equivalent relation, one can design the very constrained specific boundary gapping terms to open the mass gap of the mirror sectors.

The untouched sector in principle can provide a lattice chiral fermion model. (or, for the next step, chiral gauge theory in 1+1D.)

This post imported from StackExchange Physics at 2014-04-05 03:00 (UCT), posted by SE-user Idear