I see LFS makes use of only one core no matter the structure of the CPU be it dual,triple or quad core, I don't have the means to benchmark LFS on either of these CPU's at this moment, so, if this turns out to make any sense I'm hoping a member of the community with a triple or quad core CPU could post some benchmark screen shots with a local server hosting a full grid of AI with FPS being the markup.

This is a(very small?click on it)screen shot of Linux under load, the work seems to be in a constant state of swap with load being halved every second or so.


My problem is I cant tell under Windows if the same 'swapping' of load is true, from what I read I'm lead to believe each process in Windows has an affinity on one core of the CPU which is what I'm hoping for, Why? LFS being the CPU power hungry app it is could make use of this in a local host setup having the AI and server load set to affinity on a dedicated core whilst another instance of LFS.exe is dealing with racing and input on the other core.

example:

c:windowssystem32cmd.exe /C start /affinity 1 LFS.exe
c:windowssystem32cmd.exe /C start /affinity 2 LFS.exe(local host)
or/ I see this application for the GUI lovers SMP Seesaw

Maybe just maybe im making an ounce of sense here???

LFS does not support SMP. Yet.

Modern operating systems (such as Linux and Windows) have a load balancing algorithm for SMP. For apps like LFS that do not support SMP, 50% of each core is used, on average, although this can be less depending on total load, and the actual load on the individual cores can vary anywhere from 0-100%.

Quote from Forbin :

LFS does not support SMP. Yet.

Modern operating systems (such as Linux and Windows) have a load balancing algorithm for SMP. For apps like LFS that do not support SMP, 50% of each core is used, on average, although this can be less depending on total load.

...I don't understand why people say LFS does'nt make use of cores when clearly from what you say it uses them all rather than swap the load about - if this is the case and all cores have a load on them is it not true to say that LFS runs better because it hogs all of the cores?/maybe i've misunderstood

If we assume a single-core CPU has a total load capacity of 100%, and a dual-core CPU therefore has a total load capacity of 200%, then the max total load of a non-SMP app will always be 100%. In the case of a dual-core CPU, this load is shared among the two cores, so each core sees an average load of 50%, for a total of 100%, the equivalent of a single core.

I think that, in reality, only one core is used at a time and it just switches between them very rapidly. So yes, both cores are being used, but only because Windows is forcing it to. It's completely transparent to the app.

That`s correct. Quad Q9550 here.
Maximum load is ~25% per core, whatever i do.
Patch Z, xrg/bl1 full grid AI.