Actually, Julia's main advantage isn't pure performance. It's easy of usage. Please notice Julia is NOT a true HPC language as the only type of parallelism it provides is master-slave (as far as I know) which isn't even common in HPC. However, it's handy to be able to use Python, C and Fortran (not to mention Julia's) functions from the same environment. It's reasonably fast and provides much nicer DSLs than Python. The reason why Julia is better for this niche is that it has REPL and its JIT works pretty well. Compared to than, Nim has virtually no REPL. Ok, it has one but only for small sessions and it actually works by super-fast compiling through tcc (btw. tcc compiler backend doesn't work on my Linux machine).
On the other hand, Nim only has C and Python functions, no Fortran. Sounds not bad? It is bad, the Fortran codebase for good-quality numerical code is huge. As already said, Nim lacks true REPL, which makes it's discoverability much worse than Julia's. It also lacks in examples, tutorials and documentation. Not to mention libraries... So no, I don't thing a single "wow, it's written in Nim and it works great!" will suffice to make Nim popular.
