I am still on Organelle here, didn’t get an M yet.
But I was wondering if there actually are any patches that takes advantage of the quad core?
PD needs to launch a new version of PD for each core it wants to use. So if you want to use 2 cores, you need to launch PD two times and have two apps running at the same time.
So I was just wondering if there are any examples around that shows a meaningful way to handle that together with Organelle-M?
I know how to do it on a regular computer, but not sure how this could/would work on the Organelle. To me the ability to actually use the 3 other cores is an important feature and if I have to splash out more money, it is important to have this working.
I’ve talked about this on the main organelle-m thread a bit, so this is brief…
i don’t think anything is using multiple cores explicitly yet in PD patches, but implicitly many patches are using multiple cores a bit - because various things are done on separate threads e.g. IO.
in pure data this is done with pd~ , its pretty straightforward (and the PD help files are very clear) , but the complications tend to start when you need to communicate between the various sub-processes your creating - and how to solve this depends greatly on what your trying to do etc. (*)
Im going to be doing it for orac, and have already been considering how this might work…
Im hoping to the user, it will just appear like a multi core router where different chains are on different cores.
I’ve got a pretty good idea how to implement it, but haven’t had time to sit down and do it yet.
but pretty excited - as it’ll mean orac users will get instantly ~3-4x the processing power!
(*) you say you know how to do this on a regular computer, so i guess you know this already so this is as much for ‘completeness’ for others reading… but yes it should be working on organelle-m
Yeah I think I also have mentioned it there, but thought I’d start a thread about it.
Cool. I just think it is pretty important to be able to use more than one core for the patching. So one can make more complex patches, with bringing OG to its knees. I have some patches that barely runs on my version, so I was just thinking about if it’s worth updating.
Yes I know how it works on a regular computer and use it sometimes, to offload one processor. It does get a bit complicated at times, but sometimes it is needed
I was just not aware if something special needed to be done for the Organelle, since it is a different system. But yeah, it probably just works like on a regular computer.
I was also wondering if there would be done some special effort from C&G’s side to make it easier or simpler to work with. After all it is one of the selling points of the updated Organelle, that it has 4 processors.
Sounds good. All though I have not really used Orac yet, I prefer the old school way of patching.
For C&G to simplify it they’d need to create a framework- and before you know it, it would end up having to cover many things orac does…
Anyway I think a few simple examples would show it’s very easy to do in many cases.
one thing to be wary of , is doing things in different cores is not ‘free’ , it will always introduce some latency. ( sync audio bufs, msg overhead etc)
Yes will work, but the overhead means they are sub-optimum.
( ie they’d perform slightly worst than putting all the code into one Pd process)
Yeah,
I don’t think orac will end up as a ‘simple example’ to copy but I expect it’ll give me a very solid base to help others with Pd~ and organelle patches in the future
