deckard1 Posted May 22, 2021 Share Posted May 22, 2021 Hi, I'm using a plugin that seems to be CPU intensive...Helix Native. When I use the plugin the CPU meter jumps to around 75%, and sometimes higher, for a single core. Why doesn't Logic distribute the load to the other cores (i.e. threads) to minimize the CPU load? My buffer setting is at 128 samples...otherwise I have latency issues with recording guitar. I'm on an older MacBook Pro...not sure how much that contributes. Computer specs in my signature below. Thanks. Quote Link to comment Share on other sites More sharing options...
David Nahmani Posted May 22, 2021 Share Posted May 22, 2021 Hi, I'm using a plugin that seems to be CPU intensive...Helix Native. When I use the plugin the CPU meter jumps to around 75%, and sometimes higher, for a single core. Why doesn't Logic distribute the load to the other cores (i.e. threads) to minimize the CPU load? Because it's more efficient to keep all the calculations for a single process such as one plug-in processing one audio signal all onto one single core than to spread out over several core and then have handle that. If on the other hand the CPU load is coming from multiple plug-ins on the same channel strip then there's a way to redistribute the load onto several cores: Logic Pro/Express: Tips for balancing multi-core performance Quote Link to comment Share on other sites More sharing options...
deckard1 Posted May 22, 2021 Author Share Posted May 22, 2021 Thanks, David! Very informative. Quote Link to comment Share on other sites More sharing options...
MikeRobinson Posted December 2, 2021 Share Posted December 2, 2021 Anytime a computer program seeks to distribute an otherwise-sequential workload among "multiple processes/threads/cores," there is a certain amount of overhead which can actually become quite(!) expensive. The process must now deal with, for example, making sure that all of the various work-products delivered by the various threads are properly sequenced in time. (Much easier said than done.) And, to be sure that they do not actually waste more time, attempting to do this, than they could have gained. "Yes, I have eight trains at my disposal – but they must finally arrive, in the correct order, on one single track." This is exactly why an application might instead choose to specify that the workload has so-called "CPU affiity" to a particular, single, hardware resource. And, why this is very likely to be the best decision. The design of a "real-time application" (full discosure: "software geek" here!) very often becomes: "Hobson's Choice." Quote Link to comment Share on other sites More sharing options...
David Nahmani Posted December 31, 2021 Share Posted December 31, 2021 Facing the same issue. Does that prevent you from working in any way? If not, then it's not an issue per se, it's normal behavior. Quote Link to comment Share on other sites More sharing options...
Studio162A Posted January 1, 2022 Share Posted January 1, 2022 I use multiple iterations of Helix Native in the same project and don't experience core spikes. My iMac's specs may account for this, but you'll have to compare it to what you're using. Jim Quote Link to comment Share on other sites More sharing options...
Eon Posted January 1, 2022 Share Posted January 1, 2022 Processing audio is a simple linear math operations (add/subtract/divide/multiply) which cannot be efficiently divided to multiple cores. That's why hardware synths use special CPU's for example. It works great with linear math operations but completely suck at multi-tasking/threading operations, such as famous Motorola 56300. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.