My times with i7-3770 @4.2GHz (Linux)
Blender 2.8 = 03:02.00
Blender 2.79b = 02:23.00
And I’ve noticed that 2.8 is being quite slower than 2.79b. I clarify that the slowness is not due to the mentioned patch, it does not influence in 2.8 times without or with patch for this benchmark that does not have very many particles actually.
So I make a kind of report here to see if anyone else has noticed this with other jobs.
The patch info talks about the acceleration being in the creation of large numbers of particles, so I don’t know if it at all impacts the simulation of particles or the use of smaller numbers of particles.
I have mentioned the patch anecdotally because thanks to that I have noticed that 2.8 was being slower than 2.79b with this benchmark file. So since today the patch has been incorporated to master, I have preferred to mention it so that people who see that new commit do not think that what I am talking/reporting here could be related to that new patch/commit.
About the patch (which has nothing to do with the slowness of what I mention in this thread), you read there that it is mentioned that unexpectedly they also notice improvement in performance in baking, not just creation. Always for very long number of particles.