what sort of results do you have with this one so far?
#summer 2025 internships
lean grove
crude mulch
what sort of results do you have with this one so far?
not too much honestly -- it was just implementing a prof's paper from a while back on cpus instead of gpus, but basically boils down to a couple of operators can have an O(n^3) algorithm by using a different formulation instead of the typical O(n^4). it's pretty hard to quantify since the simulation speeds really depend on how accurate we want our simulations to be (i.e. n here)...
otherwise i'm planning a poster and a writeup, but haven't done anything on that end yet
lean grove
well i was asking about results from sending your resume out. However you gave me a better response lol; what you said is a good start to better illustrating it on a resume
sadly i don’t have much advice since i’m not remotely academic. i think there is some phds in here who might know how they displayed their research on a resume
although tbh you graduate in 3 years, i think you are easily set for internship offers?
or is research like this analogous to a industry internship for academia. don’t know how that works exactly
crude mulch
oh HAUHAUIHEFW sorry. i haven't sent it out yet!
i'm interested in some quant firms and swe jobs, honestly not too sure
i guess maybe something like this might be better?
glacial crescent
In this market just get A internship, then you can be more picky next couple of years
Did you make the Julia library? If so how did you achieve the speed up?
crude mulch
Did you make the Julia library? If so how did you achieve the speed up?
i did write the entire library, and the speedup is how long simulations took using my operators (which in practice is a specialized matrix multiplication algorithm that takes advantage of sparsity in the bernstein polynomial basis) compared to the same equivalent operators that most people usually use (which is just dense matrix multiplication). so idrk how to phrase it b/c it's just a different algorithm to do the same thing mathematically in a different polynomial basis
the catch is that the use case is fairly niche (very very high order simulations)
opal hull
remove future classes from the resume, it doesnt mean anything