Still don't see the payback in "optimizing" the BOINC program. Most research groups have limited resources (including WCG) and spending time optimizing BOINC would not be the best use of staff time. In addition, most of the science apps have been built to the "lowest common denominator" so they can be used on the greatest number and type of processors. Optimizing to the latest CPU features may actually do more harm than good to their project goals as it ultimately exclude a significant number of contributors. Once again, too much effort expended for very little, or even negative, return.

ok simple .... we optimise the boinc programs understanding of what features the user has available to utilise and what resources to report, also optimise the feature set to include more eco friendly and performance related function calls, settings.

and to report back to all boinc projects the full feature set of the utilisable hardware.

this is relatively speaking a small workload for 2 or 3 programmers.

yes and this is why we class result's by similarity,
class and geni

It's supposed to come out really soon, June or July.
Just around the corner ;)

I'm also hooked up on the Threadripper CPU, can't wait! :)