It goes both ways (in astronomy and particle physics) but I think the difference is there's more bang for the computational buck in astronomy. It takes trillion

It goes both ways (in astronomy and particle physics) but I think the difference is there's more bang for the computational buck in astronomy. It takes trillions of collisions to collect enough data to make a discovery about a single particle, for example. Interestingly, one way that computational resources have recently had an impact is with the GAIA astrometry survey mission. There's a concept in optics of an image being "diffraction limited" meaning that there is a characteristic resolution limit for a particular wavelength of light and optical instrument size. Meaning that there isn't an advantage to having pixels that are smaller than that limit, since you'll just have a blurry blob that is smeared across multiple pixels anyway, you won't actually be able to increase the resolution of what you can see. However, this isn't entirely true. What actually happens is that the light from a point-like source is spread out over an area in a characteristic pattern called the "point spread function". If you were to resolve a distant star over multiple pixels (beyond the diffraction resolution limit) then if you were able to model the point spread function to fit the data for those pixels you could potentially locate the position of the star to a precision that was higher than the diffraction limits of your optical assembly. And this is precisely what GAIA does. It uses a telescope that is much smaller than the Hubble but it has an enormous (gigapixel) CCD imager which enables it to map the point spread functions for huge numbers of stars simultaneously. Before the advent of ubiquitous high performance computing this wouldn't have been possible.