An important thing to factor in here as well is developer time. With Julia 
you can arrive at a prototype functional code in far less developer time 
than with C/Fortran. From there you can iterate on the code once you have 
identified the performance critical components. As a last step if the code 
still isn't fast enough you can write your own external C or Fortran code 
and use "ccall" just for those spots which cost the most, and you have the 
additional benefit also of having existing functional code to test against 
to see that everything still checks out.

In addition, while doing all this you have access to a full set of very 
useful tools that help to see under the hood of numerical algorithms, like 
eigenvalue or SVD calculations. You can easily interact with the data that 
julia is processing by playing with it in the REPL. 


On Tuesday, June 3, 2014 4:25:19 AM UTC-5, Thomas Moore wrote:
>
> Julia is still a very young language, and it's common to hear people 
> talking of hoped for improvements in the compiler which will further speed 
> up the language. According to the current benchmarks, Julia is consistently 
> (and incredibly!) within a factor 1.5-2 or Fortran's speed, and this is 
> consistent with my observations. I'm wondering if anyone has any idea of 
> how much we can realistically expect this to improve? Are there any 
> fundamental barriers which will prevent Julia catching up or even 
> overtaking Fortran on the raw speed front? 
>
> I ask as I'm currently doing some work in numerical simulations, where we 
> run experiments whose run-time is measured in weeks. For this work, Julia 
> running slower by a factor of 1.5 or 2 is not really tolerable.
>
> Thanks!
>

Reply via email to