How to find good universities for matter modeling?

In my experience the materials modelling field is unfortunately a bit niche to have a good ranking system. Having just done this to start my PhD, here are my tips.

Supercomputing and IT Facilities

Check out all the universities with good access to supercomputers, either operating their own on site or have a significant stake in a consortium. Not only will these resources be vital to your own PhD but good computing facilities tend to attract the best researchers. Equally, an institute with good modelling researchers will be able to get the grants to build computing facilities, so it can tell you if the place has a history of computational work. If you go and visit a school, always ask what the IT department are like and how quick/flexible they are. A bad IT department can make or break a project.

Major Contributions in Field

Have a look for people who have developed significant theories in their field that have been adopted into many other works. In my case I'd look to the people behind MD forcefields such as Goddard group at Caltech or people like Nosé, Parinello and Rahman who have developed thermostat and barostat algorithms. However, if you aren't interested in the work don't just follow a big name. Often less well known researchers have more time to discuss work with you and are more flexible to what you want to do.

Check the Type of Modelling

Following on from the point above, make sure that the modelling being done is what you're interested in. A lot of PhD adverts state "modelling" as a buzzword without specifying what simulation scale or focus. From my own experience I'm interested in organic polymers, but this often comes under "soft matter" which also includes lots of biological/drug modelling work. Many people want to start incorporating modelling into their work, but a lot don't have the background and experience to help a PhD student get started in the field.

Best of luck finding a PhD!