Is there anyone who's sufficiently qualified in astronomy/astrophysics who could point out what advances in HPC/modelling/signal processing will be necessary to

Is there anyone who's sufficiently qualified in astronomy/astrophysics who could point out what advances in HPC/modelling/signal processing will be necessary to accelerate things like this? I don't know telescopes but I know enough about DSP to make me suspect this is gonna need all sorts of fun high performance deconvolution algorithms....