[LAU] i5 Hyper-Threading, BIOS settings and Arch n00b pointers

[dale]

On Thu, 2014-08-28 at 22:40 -0700, Len Ovens wrote:

> Performance is an interesting word, hyperthreading has always helped 
> performance, even audio performance if you are talking about CPU 
> throughput. In jackd terms, even on the single core P4, audio was pretty 
> solid down to about -p64, then xruns. With hyperthreading off -p16 with no 
> xruns was possible. However, overall performance was less too.
> 
> > So can anybody point to any conclusive evidence that i-series processors benefit
> > from having HT disabled on a Linux based DAW? Preferably benchmarks on a system
> > installed with HT Enabled and Disabled using a recent kernel and system.
> 
> That does sound interesting. There is a site that tests lowlatency 
> performance of a number of machines with RT kernels. Here is one for an i3 
> with hyperthreading:
> https://www.osadl.org/Profile-of-system-in-rack-0-slot-0.qa-profile-r0s0.0.html
> 
> It looks very good, so your comment about the difference between old 
> hyperthreading and new may be right. (the highest latency in the plot is 
> 34us with cycle test) How these test relate to real world audio use I 
> don't know. At the top of the above page there is access to the rest of 
> the rack slots with different machines that might be there.

Have done some further reading, and although I can't say for definite it
sounds like support for modern hyperthreading mode is quite well
supported in Linux now and for general/desktop use. Seen one report that
it specifically caused problems with (some?) rt-kernels, but not with
generic. No mention of pre-empt. Another reporting that it increases
power but generally at the cost of latency.

I would be interested to know if it would allow you to play something in
your DAW that uses greater CPU power but at the cost of low loads
requiring a higher buffer setting within Jack. It does seem these days
it may be down to the question of whether minimum latency or maximum
available power is of most importance to you. (Although without some
real checks I'm only just postulating. If somebody wants to provide me
with a project file that will play on a default install of KXStudio
installed with it turned on and off.

Although that still begs the question of whether it being turned off in
BIOS/UFEI before or after installation actually makes any difference...

> 
> > CPU Management
> 
> Not sure what you mean by that. SMIs? Any time the OS doesn't know what 
> the CPU is doing, RT latency suffers. Intel agrees turning them off 
> improves low latency, but they also say that in some cases this voids 
> warranty...

I now think this is CPU Power Management and turns off some levels such
as Sleep or Hibernate but I'm not positive...

Thanks, Dale.