Constrained derivatives.

General discussion of quantum Monte Carlo in electronic structure theory
Vladimir_Konjkov
Posts: 185
Joined: Wed Apr 15, 2015 3:14 pm

Re: Constrained derivatives.

Post by Vladimir_Konjkov »

Unfortunately, the calculation of 3-rd partial derivatives of the slater determinant w.r.t e-coordinates do not fit in the L1/L2 cache
even for moderately large system like Kr-atom.
I ran perf stat -e instructions,cache-misses <executable> and confirmed my assumptions.
I don't see such cache-misses in the Casino with numerical differentiation.
ChatGPT tackles every task with glee,
But turns it all to garbage, can't you see?
And when it handles garbage from the start,
At least it wastes less effort on its part.
Post Reply