While many of these optimization may have a place in a computer being use for a DAW, it does not follow that they also have a place in a computer being used for listening to music.
A DAW has some important requirements:
- Low latency throughput from input to output (audio interface input to audio interface output)
- High volume of audio data processed by heavy use of floating point calculations in real time
- High volume of audio data pulled from storage and processed (including being mixed on a summing bus).
- Sometimes high volume of data being written to storage
- All of above with very tight timing being maintained with no disruption of any ADC or DAC data paths.
Additionally, a decade or more ago many of the plugins and even some DAWs were not good at multi-CPU synchronisation.
Disabling hyperthreading back in this time was though to assist in making the floating point parts of the CPU more consistently available to an FP intensive high priority DSP thread. A read up on exactly how hyperthreading used to work (and maybe still does?) will probably help, but simplistically, a single CPU has one of a bunch of important functional units including floating point and integer units. Hyperthreading allows two virtual cores to each bit of the shared core that another virtual core is not using. So, it doesnt for eg increase floating point performance but did allow integer and floating point to occur concurrently per core.
I have experimented with such tweaks back in those days being a semi-professional producer at the time. In end - yes sometimes you could get a benefit, but often I found it hit and miss according to the specific intel hardware and memory etc in use and even OS. Either way, I generally found if there was a gain it was marginal and as DAWs improved then actually the reverse was more often true - ie disabling HT was much slower. DAW are typically not fine grain multi-threaded apps as the overhead of thread sync are quite high, along with CPU caching strategies of the time etc - it made some sense that this use case could benefit from HT off.
The other thing that is mentioned however did hold true back at that time - disabling power related clock throttling (speed step etc). The main effect of disabling this was to allow the DAW to be pushed to higher overall utilization before audio drop outs occurred because at the time it took too long for the CPU clock to speed up again to avoid an audio drop out under very low latency conditions.
Now bare in mind, I was working with very low audio buffer size in order to achieve the lowest round tri latency possible (ie latency between me playing a key on my midi keyboard to a software synthesizer making a sound, to me hearing the result via an audio interface). This was important as too much latency (more than a few ms) making maintaining accurate timing (groove) when playing an instrument quite hard.
NONE of these issues apply to music playback of even high res on most modern computers.
Also clock throttling for power has become vastly more responsive these days compared to when such tweaks were initially used. These day I do not apply any such tweaks on my windows 10 PC running ableton live. I had never need to apply such tweaks on my macs at all as they were always able to maintain glitch free audio under even very heavy load and these days windows 10 has become a lot better as well. Also modern DAWs are much better implemented in terms of how the perform cross thread synchronisation with multiple cores. Back then when these kind of guidelines were written they were often poor.
None of these tweak were ever done for audio quality reasons in a DAW - they were done very specifically to get the most out of the available hardware and software of the time.
Even at the time (early 2010s) power throttling (speed step etc) was never an issue for typical normal/high latency audio associated with listening to music that typically uses very conservatively long buffers at multiple stages because quite simply there is no hard real time requirement.
These days compiler have very much improved. The hardware implementation of HT in the CPU has improved and the OSs have significantly improved. Switching this off is more likely to cause issues in a marginally capable system than prevent them.
Also disabling any form of power related clock throttling will result in the CPU running at a fixed and often max speed regardless of load. This mean it will run hotter and consume much power (so the power components also run hotter). Unless very well ventilated this may well cause way more serious problem including complete thermal shutdown which on many motherboard also results in a full BIOS reset for safety reasons.
