Julia is inspired by Lisp, Python, R, Fortran and C/C++. It is aimed at highly scalable performant numerical and statistical computing while being easy to use in a similar way as Python.
Its execution speed in many instances is equivalent with C, C++ or Fortran.
Intel Labs is a heavy user of Julia. They develop the
ParticleAccelorator.jl library which adds an additional layer of optimization on top of Julia's parallelization methods resulting.
Julia offers a performance similar to C while being higher level, similar to Python or Matlab.