CPU usage Maschine on a 2014 MBP compared to 2017 iMac

I started using my mid 2014 MBP again after using my 2017 iMac since the end of 2017. Both computers are fully SSd'ed with 16gb Ram. They are almost identical specification wise. I bought the iMac due to constant CPU issues with NI products. The iMac is running way more efficiently despite being almost the same specification wise.

For example, when i open the same project on both computers the CPU usage on the MBP is more then double that of the iMac. I narrowed the culprit down to Maschine. With one or 2 groups open my CPU would drop from 50-60% resulting in sound dropouts and crackles. After I deleted that instance of Maschine it'd go to 5 - 10%. The same project on the iMac wouldn't even be pushing 20% with Maschine running 3 groups! This pattern is the same with all my Live Set's.

I have spent so much money on new faster SSD's and better Soundcards hoping this would remedy the situation with the Macbook Pro (mid 2014 15" Retina) and the Maschine issue. No luck it seems.

My Mac OS is updated. All NI products are updated and no apps are running in the background. There is simply something with this MBP and Maschine that do not align. Ideas?

 

What is you processor on the MacBook pro and Imac?

 

MBP = 2.5 ghz quadcore i7

iMac = 3.5 ghz quadcore i5

Both machines should eat a single instance of Maschine with 2 - 3 groups running. No problem!

 

Well for starters your Imac has 40% more power. Not quite the same specs.

 

I have a dual core 2.6ghz from 2009 on my Imac and it handles up to 8 groups no problem

 

I am sorry but CPU power at this level should not be noticeable. If I was rendering 4k video I would expect to see it. As the poster below states. His 2009 model has no problem.

 

4k video is handle by your graphic card mostly... Single core performance is much higher on your Imac same goes with cache memory they are actually not comparable specs

 

I am sorry Kaldosh. I understand what you are trying to say but I have friends with computers 50% less powerful than mine and they open the same sets with 50% reduction on cpu compared to mine. Something is wrong with this box. I have upgraded my soundcard to one of the best and most stable products ever made (RME Fireface UC). I use an external T5 Samsung SSD running at a tested 460mps read/write. Fully loaded RAM and a factory installed 500gb SSD. Anyone? Is it this motherboard? I am starting to feel this is hardware related.

 

No it is not, I don't think so... Are you using an antivirus?

 

Nope. No anti-virus.

 

It's most likely because the 2017 MBP has a newer generation CPU compared to the iMac.

 

Right. Your description states 2014 Mbp and 2017 Imac Is it the other way around?

 

It is mid 2014 MBP - 2017 5k iMac. In theory the MBP should keep up with it. An i7 versus i5 procesor (i7 in the MBP!)

 

The frequency is much lower but a quad core from 2014 should indeed keep up no problem with it... Maybe you just upgraded your OS several time and a clean install would actually fix the issue. Also if the hard drive is pretty full performances slow down...

 

That depends on the software actually. On OSX Adobe (Premier, After FX) tend to use more CPU than GPU while Final Cut or Davinci Resolve uses both but mostly the GPU. Video on Mac is kind of complicated due to software/GPU optimization.

I might boil down to the difference in CPU generation, one is a 4th gen CPU and the other is a 7th gen, that's quite a gap. The Clock on the iMac is indeed 40% faster. According to synthetic benchmarks (geekbench) the iMac is 25% faster for single core performance and but has negligible difference in multi-core performance. (synthetic benchmarks don't always equal real-life performance though)
Regardless of all that 60% usage with only 2 Groups is way too much, something is wrong.

Can you open the "Anti Blocker" project on your MBP and tell me how much CPU it uses? It's on the default library and has 7 Groups, on my 2015MBP (Dual-Core i5 2.7GHz) CPU usage is around 25% with a 512 Buffer Size (Activity Monitor reports 85% but that's not accurate as it can go up to 200% or more...)

This is the MacBook Pro, right? MacBook Pro i7 2.5GHz 15" Mid-2014, (I7-4870HQ - 4th Gen CPU)
And this is the iMac, right? iMac Core i5 3.5GHz 27-Inch (5K, Mid-2017) (I5-7600 - 7th Gen CPU)

 

I have tried them all. Thanks for your words. Always cool.

 

Thanks so much D-One.

Yes, those configurations you wrote last are correct for the 2 machines in question.

I just opened "anti-blocker" in Ableton with Maschine 2 (VST) running as a single instance in channel one. CPU runs at 23%. That seems to be normal(ish) yeah?

 

My first thought would be the same as D-One's – different CPU generations, but...

My trusty 2011 MBP (2.2GHz Early-2011 i7-2720QM) had discreet GPU failure recently so I needed to get something *new*. I got 2015 MBP (2.8GHz Mid-2015 I7-4980HQ) which is almost identical to OPs from 2014 (and would be the same if I had 2.5GHz version). To make the story short – thanks to internet knowledge (long live world wide web!) I have revived my 2011 MBP (switching off discreet GPU) and now have two MacBooks to compare.

So, first of all – my MBP2015 works fine with lots of groups as was my MBP2011. So I just opened Anti-bloker on both machines and they use almost the same amount of CPU. I know D-One said that Activity Monitor is not accurate, but jus as a reference MBP2015 uses 85% and MBP2011 uses 99%. So not that much of a difference.

But then I realised OP is using Maschine inside of Ableton Live. I'm not a Live user, but I got Live 10 demo, so my results were quite surprising. MBP2015 uses 23% (shown by Live CPU meter) and MBP2011 uses almost twice as much – it goes up to 40%. In Activity Monitor Live uses 71% in MBP2015 and 112% in MBP2011.
So maybe Live 10 is not optimised for use with older CPUs?

But I also thought that maybe Maschine plug-in caused problems so I did the same test for Bitwig 2 (with Maschine as a plug-in) and results for both laptops were pretty similar for both computers. What was surprising MBP2011 was using a bit less CPU then the new one. Activity Monitor was showing 66% on MBP2011 and 78% on MBP2015.

My both macs has the same system (actually one is a copy of another) – OSX 10.13.6, both have 16GB of RAM and SSDs. I always do clean installs on my computers.

Have you tried using Maschine standalone and comparing them?

 

Thanks for the comparison there Jiloo. It turns out my figures are not so unusually high. Certain Maschine Groups are more intensive then others (depends on what instruments / effects they contain I guess).

FYI in standalone mode I ran "Ani Blocker" and it is hard to gauge because I don't see a CPU numerical value anywhere but the meter itself wasn't even registering. That means Ableton and Maschine (in fact most NI) products do not work so well. We all know that NI and Ableton are not friends but I'd really like to know who is responsible here for making their products so CPU in efficient in Ableton? Is it Ableton holding back on code or is NI simply too lazy to address the problem?

 

this is a thread about mac's, but i tried "anti-blocker" on my old (second-hand) laptop, with i7-2630QM, 4GB, and 180gb ssd (built in by me...) and windows 10 (1803).
it is not the computer i use for "real" music production, i use it mainly for learning synths, making sounds, play around...

it doesn't goes beyond 20% (task-manager), it looks like the average load is about 16-18%.

i did use the wasapi driver.

with asio for all, 512 samples, it stayed most of the times at 15%, with peaks of 18%.

i used the built-in soundcard, a realtek something. not a really good one..

the turbo boost of the i7 is by the way 2.6 ghz. it is an old one... (i didn't disable background services, like virusscanner, etc.)

perhaps interesting... the difference between windows 10 and mac os.

EDIT: when checking process control, (i am dutch, with a dutch windows, normally i was used to english), maschine 2 reports as highest 10%, where the other 5-7% comes from i couldn't detect. even services report 0%, strange.... perhaps it doesn't read that well, the real performance...

EDIT 2: maschine 2.7.7 i used (yoda talks...), it reports that it uses SSE, SSE2, SSE3 and AVX.
ánd maschine uses (forced) the geforce gt540m, not the built-in graphics card, that makes a big difference.