EBQM Research Log 4

Diary of a Marie Skłodowska-Curie Research Fellow

One of the promising ideas that had to be discarded this month…!

Diary of a Marie Skłodowska-Curie Research Fellow: Entry #4

Hello, and welcome to the fourth post in this series of Research Logs for the Marie Skłodowska-Curie project ‘Ergodicity Breaking in Quantum Matter’ (ebqm.info). You can find the first post here.

The frantic pace continues! It’s been another busy month – so busy, in fact, that I forgot to write this post in time for the first day of the month as usual, and so I’m currently putting this together at 11pm on the second as it’s the first bit of free time I’ve had this week for writing! It’s been a month of ups and downs, with several seemingly promising research ideas coming face to face with a far less promising reality, but there is a certain liberation in being able to escape the shackles of an idea doomed to failure and instead take what’s been learned and move on to other ideas that do work. At least, that’s what I’m telling myself. ;)

Somehat incredibly, this month marks the 25% mark of the Ergodicity Breaking in Quantum Matter project, and consequently the same percentage of my time in Berlin. Between the pandemic and the winter months, I feel like I’ve barely scratched the surface of the city and I’m looking forward to the spring and summer to try and make the most of my remaining time here. On the scientific side, I’m starting to feel the pressure of not having published anything from the EBQM project yet, but I have quite a few things in the pipeline and I’m not willing to rush anything out until it’s as good as it can be.

That said, with the realisation that a quarter of my time here has passed, I’m already having to think about next steps and what I want to do next. For the moment, I’m still holding on to the dream of starting my own research group, but with the academic landscape rapidly shifting in many ways for the worse, I’ve been giving a lot of thought to how I can best be a positive force for good in the community while also doing the best science that I can, and securing the best possible opportunities for any students who join me on the journey. I don’t have any answers or specific plans yet, but I know that if I’m to take another step up the academic career ladder, it has to be in the right environment and for the right reasons. Over the next few months, I’m going to engage a bit more with the ‘professional development’ aspects of the MSCA programme to try to figure out how best to develop the skills I’ll need for whatever my next step may be.

This has been the second month of my “View from the arXiv” series of weekly posts highlighting newly released papers on the arXiv preprint repository, which so far have done reasonably well. I’ve enjoyed the process of writing them, and the feedback I’ve received from others has been positive, so they’ll continue for at least the near future. Likewise, I’ve also been posting popular science blogs over at Broken Symmetry which have also been well-received, and I’ve enjoyed getting back into writing. I’ve enjoyed it so much, in fact, that I have plans for another popular science venture, one that will require significantly more funding and support in order to make it happen - more on that if it materialises!

One of this month’s highlights was the theory discussion seminar I gave at the University of St Andrews on the topic of flow equation techniques for the investigation of many-body localized quantum systems. It was fantastic to (virtually) return to St Andrews, after not having been there in person in 3-4 years, although it was a little sad that many of the staff members who I knew from my time there couldn’t make it due to the ongoing UCU strike action. That said, I fully support my UK colleagues in their strike action and wish them all the best in the ongoing negotiations, as the current pension cuts seem cruel and callous in light of the working conditions of the last few years. (I am also extremely thankful to one lecturer who was there and asked a few questions that weighed on my mind and might’ve just led me to the solution to a problem I’ve been wrestling with for months - if this pans out, I’ll be sure to drop him a mail to thank him!)

Work has continued on my numerical library for the flow equation codebase, now tentatively titled PyFlow. I’m planning a public release in the second half of the year, but could possibily release a less polished alpha version much sooner in order to gauge interest and gather feedback. The thing is, the library currently contains a bunch of unpublished innovations, so while I’m tidying up and smoothing out the code, I’m also trying to finish up two two main papers detailing the new features of the library. I’ve been experimenting a bit with various numerical algorithms to speed up the code, and interestingly have found that while NumPy’s tensordot doesn’t parallelise nearly as well as I’d hoped on our cluster and Numba’s jit method parallelises nicely but doesn’t achieve a particularly great return in terms of the numbers of CPUs used, somehow the single-core performance of Numba’s vectorisation method is only ~30% slower than the other methods running 20 CPUs, meaning I could instead use all those CPUs to run different simulations, representing a vast speed increase over what I could manage previously. That said, to get to the really exciting stuff, I still need as many cores as possible - keep your fingers crossed for good news regarding my NVIDIA grant application! (As I write this, I should find out the verdict tomorrow…!)

I’ve also been doing quite a bit of work unrelated to flow equation methods, as some interesting possibilities have arisen using existing techniques. Much as I think flow equations are a fantastically useful tool, developing this stuff takes a lot of time and I’m only one person - it’s a lot easier to make fast progress using existing methods, if you can find an interesting problem to work on, and this month I think I’ve spotted a few novel research directions that no one else seems to have hit on yet. Developing those will be a background activity over the next month, while trying to finish up some existing projects.

And although not strictly EBQM grant-related news, I’d like to end this month’s log by highlighting a new preprint that I worked on with my colleagues at the Ecole Polytechnique, just south of Paris, titled ‘Entanglement spectrum and quantum phase diagram of the long-range XXZ chain’. Please do check it out if you’re interested in novel entanglement measures!

Dr Steven J. Thomson
Dr Steven J. Thomson
Research Scientist at IBM Research

Theoretical condensed matter physicist, currently a Research Scientist at IBM Research UK.

Related