I should visit JET sometime. Of all the systems on the machine I worked on (https://hsx.wisc.edu/), the coil current feeds were the most difficult. Trying to cram that much current through a small piece of copper with a discontinuity takes years of effort.