How's the CPU efficiency in Live 10?
Re: How's the CPU efficiency in Live 10?
had a problem with cpu usage managed to fix it
Last edited by ash1 on Tue Nov 07, 2017 11:50 am, edited 1 time in total.
Re: How's the CPU efficiency in Live 10?
Actually it's not early stage at all, so if you feel something's wrong please report on Centercode with details/specs and status report.ash1 wrote:so far the ableton beta 10 cpu usage is bad before you even load an effect the meter is on the move
its early stages so im sure it will get fixed as we go along
Ableton Forum Moderator
Re: How's the CPU efficiency in Live 10?
Is it down?[jur] wrote:...please report on Centercode with details/specs and status report.
Re: How's the CPU efficiency in Live 10?
NVM: it's back...
Re: How's the CPU efficiency in Live 10?
managed to fix my cpu usage problem by updating sound drivers
Re: How's the CPU efficiency in Live 10?
ash1 wrote:managed to fix my cpu usage problem by updating sound drivers
Re: How's the CPU efficiency in Live 10?
I recently used GCC from Developer Tools on MacOS Sierra with "-j 5" argument, which allowed to compile a large open source project and allocate jobs to available cores. While compiling, the CPU was at it's maximum, according to activity monitor. I thought the technology behind thread-based processes should be the same, although maybe it's a lot more complex when it comes to processing 3rd party plugins. Another MacOS application is HandBrake, which can encode (render!) videos and utilise all CPU cores at the same time. Anyway, I thought I would provide an example where it works.Stromkraft wrote:I think for this to be possible there would need to be a breakthrough in how software interfaces with modern CPUs, so that cores can be presented to the software as one core and used that way effectively, also when a code thread has started to work on something. I'm outta my depth on how exactly these things work now so my take here is conceptual.locojohn wrote:
Suppose there's a heavy effect on a particular track. Why can't Live utilise ALL cores to process this effect if it detects there's less work for other cores at that moment? Eg why should all other 3 cores be stuck while a single core is trying to complete a CPU-intensive operation?
I simply assumed Live 10 could handle this differently, this is why I asked about rendering performance.
The OS can distribute work to some extent, but this functionality may not be up to the needs of the music software in this case, because a render are made by so many inter-dependent calculations and handing over work to threads running on another core does not come free, especially as likely results need to come back when ready. So there's a resource cost involved.
Andrejs
/*
- the basic tone of life remains the same,
and in it there are some happy melodies
and some sad melodies
- sekito kisen
-
- Posts: 7033
- Joined: Wed Jun 25, 2014 11:34 am
Re: How's the CPU efficiency in Live 10?
Yeah, but none of your examples involves interdependent calculated audio streams (that is not ready-made simple bit streams, but a seemingly endless line of nested calculations creating one such stream per each routing and per each processing stage and each summing stage) worked into a time domain dependent finished product.locojohn wrote:
I recently used GCC from Developer Tools on MacOS Sierra with "-j 5" argument, which allowed to compile a large open source project and allocate jobs to available cores. While compiling, the CPU was at it's maximum, according to activity monitor. I thought the technology behind thread-based processes should be the same, although maybe it's a lot more complex when it comes to processing 3rd party plugins. Another MacOS application is HandBrake, which can encode (render!) videos and utilise all CPU cores at the same time. Anyway, I thought I would provide an example where it works.
Andrejs
The structure of a binary app is not time-domain related in its nature. It has a completely different structure. There's probably some interdependence during build, but not at every single step as most apps are modular, parts are likely built separately.
Handbrake calculates something that is already known, pre-existing video frames in an already defined order and some pre-existing audio streams. The interdependence is very small compared to a song in a DAW.
I'd love to learn more about these things, certainly those that are on the forefront of creating new paradigms.
Make some music!
Re: How's the CPU efficiency in Live 10?
Yep, it's a different thing, as I'm sure you already know.locojohn wrote:Anyway, I thought I would provide an example where it works.
Anyway, the simplest way to think about the fundamental problem in audio processing parallelization: imagine you have something like eight very resource-intensive plugins inserted on one channel. They max out the CPU in your computer and you check the OS CPU meter, and find that there is plenty of capacity left across the cores (say, eight of them). The thing is, it is impossible to distribute the load so that each of those plugins is processed at the same time, each on a different core. The first plugin has to reach a result that it passes to the next one in the chain, and so forth. If you calculated all of those plugin processes at the same time, there would need to be some paranormal crap going on as most of the cores would be working on data that doesn't exist yet.
Now, further imagine those eight separate plugins are actually the different stages of a super heavy single plugin. The audible outcome is the same, as is the internal audio processing logic inside the plugin. The DAW doesn't "see" what is going on inside that plugin, and it has very little say in trying to break that process into simultaneously calculated pieces on different cores, as it's impossible to do so even if the stages of the process were discrete plugins inserted on that same channel.
-
- Posts: 4478
- Joined: Wed Oct 24, 2007 4:50 am
Re: How's the CPU efficiency in Live 10?
actually on the latest hardware, realtime processing is harder to achieve, even though there is theoretically more overall processing power available.
one problem is that fancy low-level power management can cause glitches that applications have no control over. and, as others have mentioned above, the coordinating and scheduling of multiple streams of processing is actually very tricky business.
in fact, if we could develop much, much more powerful processors then there would be no need to even mess with multicore, hyperthreaded machines. using multiple cores and parallel processing is more or less a way to overcome the limitations of single-core processing power. yes, using more cores provides more overall power, but at the cost of a lot of added complexity.
one problem is that fancy low-level power management can cause glitches that applications have no control over. and, as others have mentioned above, the coordinating and scheduling of multiple streams of processing is actually very tricky business.
in fact, if we could develop much, much more powerful processors then there would be no need to even mess with multicore, hyperthreaded machines. using multiple cores and parallel processing is more or less a way to overcome the limitations of single-core processing power. yes, using more cores provides more overall power, but at the cost of a lot of added complexity.
-
- Posts: 405
- Joined: Tue Aug 12, 2008 4:57 am
- Contact:
Re: How's the CPU efficiency in Live 10?
For me on my late 2011 mbp the cpu use is way better!!! It's also more stable for me.
-
- Posts: 7033
- Joined: Wed Jun 25, 2014 11:34 am
Re: How's the CPU efficiency in Live 10?
My initial report using Live 10b146 on my backup machine, a Core2Duo MBP using RME Babyface, that yesterday struggled in Live 9 on a 1024 samples buffer, occasionally crackling at around 80% engine load.
Live 10b146 running the same project is now not crackling even if the load engine meter goes up to 90%, which makes this meter seem more accurate than before. This is at a 512 samples buffer and running background apps, like Firefox with many windows open, that were not present yesterday with Live 9.
The project contains the MFL audio effect Magnetic.
This is just one quick impression, but it's looking good. Chances are there have been some smaller improvements in the performance area.
Future will tell if this affect also other projects.
Live 10b146 running the same project is now not crackling even if the load engine meter goes up to 90%, which makes this meter seem more accurate than before. This is at a 512 samples buffer and running background apps, like Firefox with many windows open, that were not present yesterday with Live 9.
The project contains the MFL audio effect Magnetic.
This is just one quick impression, but it's looking good. Chances are there have been some smaller improvements in the performance area.
Future will tell if this affect also other projects.
Make some music!
-
- Posts: 588
- Joined: Thu Jan 08, 2015 12:37 pm
Re: How's the CPU efficiency in Live 10?
...that sounds promising!!!
macbook pro m1pro, macos monterey, rme multiface via sonnet echo express se I, push 2, faderfox mx12, xone:k2
Re: How's the CPU efficiency in Live 10?
Complex and Complex Pro Warp Modes use substantially less CPU here, seems like around 40% less.
♥♥♥
Re: How's the CPU efficiency in Live 10?
Indeed, some really really promising experiences, thanks!