Interesting question Misha. I never tried to time this before, but this machine is faster than my other two.

Here's how I set this experiment up. I selected 11 previous songs at random that I have worked on. I unfroze the styles that I had in the original song and selected a new style at random. Hit the stopwatch and "Generate and Play" button at the same time and ran the time until I heard music. I typically add RT solos to my 3 verse songs on the entire 2 chorus, in those songs with a RT solo regeneration seems to take longer.

Here are the results, with number of RT's listed first, then the regen time in seconds.

3 - 2.9
5 - 5.2
5 - 7.2
5 - 5.5
5 - 11.2
5 - 5.6
5 - 4.7
6 - 7.2
6 - 6.8
6 - 8.7
6 - 9.2

I really don't know how this compares to anyone else's regen times. It's definitely faster than my other 2 Intel i5 systems. Those systems both had programs on SSD's and secondary 7200 rpm HD's for my RT's with 12 and 16GB of memory respectively. There is a difference between the number of cores and threads on i5 versus i7 but I'm not clear on how BIAB handles those.

As others have stated, overclocking is not for the faint of heart. It will definitely void any warranty that you my have, your system will run hotter and at higher voltage; likely shortening the computer's life. You really have to run very long stress tests to be sure that your system is stable and can handle this playing around. I'll typically run stress tests for 24 hours straight or more.

In my case, I'm a technician and came into music because of the technology not the other way around. I'm sure that explains my tendency to tinker and my lack of music theory which many folks here are expert at.

Jeff

Win11, Intel i7 7700K 4.2Ghz, 32Gb RAM, 2x1Tb HD, 500Gb NVMe, BIAB/RB 2024, MOTU 828MK3 audio, MOTU Midi Express, Yamaha Montage 7, DX7II, TX802, Motif XS Rack, Roland Fantom XR Rack, Oberheim Matrix 1000, VoiceLive3 Extreme, Kontakt 6, SampleTank 4.3