« first day (412 days earlier)      last day (1041 days later) » 

2:24 PM
I'm currently writing a thing about accelerating things with better AES implementations. Given my previous answer (mattermodeling.stackexchange.com/a/1682/1128) it seems matter modelling people do a lot of randomized simulations requiring lots of random numbers. Do you guys by chance have a default paper to reference for such things?
(i.e. a paper or set of papers doing fancy things with random numbers and potentially even having random number generation as a serious performance concern)
paging @Camps as a mod and asker of the question 👀
 
 
3 hours later…
5:43 PM
@SEJPM I'd be happy to help you with that. We do a lot of Monte Carlo simulations, which require random numbers. There's so many applications of Monte Carlo in matter modeling. One paper from the 1950s has 45000+ citations on Google Scholar: aip.scitation.org/doi/10.1063/1.1699114
Now while the title of that paper says it involves "fast computing machines", it was the 1950s and HPC (high performance computing) was different then that it is now. If you want a more recent reference, this is a paper in which Monte Carlo is used for a matter modeling problem in a truly HPC sense (more than 24000 CPU cores were used!): aip.scitation.org/doi/full/10.1063/5.0005754
Even though these papers were almost 70 years apart, the were both published in the same journal (Journal of Chemical Physics)! If you want more classic papers on the topic, I'd be happy to help, you can email me at nike@hpqc.org
 
6:06 PM
@NikeDattani these should work :) Thanks!
 
6:21 PM
@SEJPM Great!
 

« first day (412 days earlier)      last day (1041 days later) »