Macbook pro vs Mac pro, my Macbook pro is faster??

Discussion of anything not related to audio or music production
Post Reply
theinfinate
Posts: 196
Joined: Sat Jan 22, 2011 10:21 am

Macbook pro vs Mac pro, my Macbook pro is faster??

Post by theinfinate » Sat Aug 20, 2016 8:59 am

I've been saving for a while now to buy a mac pro as my macbook pro (i7 2.6 quad) suffers recording at 96 kHz and having multiple instances of omnisphere, contact, addictive drums etc. I just received my mac pro 2009 2x 6 core 3.33 ghz thinking it would do the trick. I don't know if i haven't configured something correctly but with 3 instances of omnisphere open on both computers, macbook pro is sitting at 15 percent and the mac pro is sitting at 22 percent. Can someone she some light on this situation?

Stromkraft
Posts: 7033
Joined: Wed Jun 25, 2014 11:34 am

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by Stromkraft » Sat Aug 20, 2016 11:45 am

theinfinate wrote:I've been saving for a while now to buy a mac pro as my macbook pro (i7 2.6 quad) suffers recording at 96 kHz and having multiple instances of omnisphere, contact, addictive drums etc. I just received my mac pro 2009 2x 6 core 3.33 ghz thinking it would do the trick. I don't know if i haven't configured something correctly but with 3 instances of omnisphere open on both computers, macbook pro is sitting at 15 percent and the mac pro is sitting at 22 percent. Can someone she some light on this situation?

Generally the Mac Pro model I think you got — You're not very detailed, but must be the 2010 model no?  —  is 18% faster in multicore Geekbench tests but about 15% slower in single core Geekbench tests. Is it the 2012 MBP you're comparing with?

One of the first things to check anyway is if toggling multicore support in your synths make a difference or not.
Make some music!

theinfinate
Posts: 196
Joined: Sat Jan 22, 2011 10:21 am

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by theinfinate » Sat Aug 20, 2016 12:44 pm

Generally the Mac Pro model I think you got — You're not very detailed, but must be the 2010 model no?  —  is 18% faster in multicore Geekbench tests but about 15% slower in single core Geekbench tests. Is it the 2012 MBP you're comparing with?

One of the first things to check anyway is if toggling multicore support in your synths make a difference or not.[/quote]

Ah sorry i should have mentioned i's a 2009 mac pro with updated firmware compared to a 2012 macbook pro. Thanks, il try toggling the multicore support tomorrow


hacktheplanet
Posts: 2846
Joined: Sat Sep 25, 2004 6:37 am
Location: Chicago, IL
Contact:

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by hacktheplanet » Sat Aug 20, 2016 1:51 pm

This seems to indicate that the CPU inside the MBP will have better performance: http://cpuboss.com/cpus/Intel-Xeon-W356 ... -i7-3720QM
These days, there's a lot more to consider than frequency when it comes to CPU power.

I bet enabling multicore support will improve performance since the MP has more CPU cores.
Image

Angstrom
Posts: 14678
Joined: Mon Oct 04, 2004 2:22 pm
Contact:

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by Angstrom » Sat Aug 20, 2016 2:38 pm

Heh, that's pretty much what my deleted post said. Then I realised "Christ, how depressing, this guy saved up for a machine older and less powerful than the one he had anyway" . So I deleted it!

Well, in addition, in late 2009 there was a fad for owners upgrading firmware and adding those 6core x5500s to the under-nourished early 2009 powermacs. so if he has one of those it wouldn't be a great sign.

But thats double depressing.
So, there you go. Depression for everyone!

theinfinate
Posts: 196
Joined: Sat Jan 22, 2011 10:21 am

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by theinfinate » Sat Aug 20, 2016 10:32 pm

Multicore support was on and doesn't really make much difference anyway. So the guy i bought it off told me lies, i told him what i already had and if it would be a great improvement and he said it would be a lot faster. Is this possible, i'm so pissed off spending 3K on a computer that hasn't enhanced my studio at all, wtf aarrrrrrrrrrgggggggg

beats me
Posts: 23319
Joined: Fri Mar 30, 2007 6:39 pm

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by beats me » Sun Aug 21, 2016 12:31 am

8O 3k would get you a brand new 27" retina iMac almost fully decked out with all the options.

That might be an Angstrom "Did I just say that out loud?" post, but did you look into that option?

theinfinate
Posts: 196
Joined: Sat Jan 22, 2011 10:21 am

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by theinfinate » Sun Aug 21, 2016 6:58 am

beats me wrote:8O 3k would get you a brand new 27" retina iMac almost fully decked out with all the options.

That might be an Angstrom "Did I just say that out loud?" post, but did you look into that option?
No I trusted a mate who I thought knew his shit when he steered me towards getting a Mac Pro tower as it can be customised, I thought it was simple maths behind the processing power being faster as in 4 x 2.6 vs 12 x 3.3, I thought that was a no brainer??

Stromkraft
Posts: 7033
Joined: Wed Jun 25, 2014 11:34 am

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by Stromkraft » Sun Aug 21, 2016 7:56 am

theinfinate wrote: a Mac Pro tower as it can be customised, I thought it was simple maths behind the processing power being faster as in 4 x 2.6 vs 12 x 3.3, I thought that was a no brainer??
The processor speed is not as interesting as how much machine code can be executed in how many threads in a given cycle or specific time frame, which is why bench scores and machine comparisons should be looked up before making a purchase. In theory, 12 real cores should be able to outperform 4 cores even if a little slower, provided one core isn't overloaded and all cores can actually be utilized. There are also generational differences between processors. A 2009 CPU is very old (7 years!) in the computer world.

Can you for instance do this test or similar:
  • Set your instruments in single core mode if they have such a setting.
  • Create as many similar instrument tracks as you have real cores — not counting virtual — playing a longer section, like 32 bars, with 16ths. Make the tracks somewhat different to each other in the MIDI events.
  • Set the buffer size so that playback is smooth.
  • Open up Activity Monitor and press Cmd-2 and look over how the cores are used.

Are the cores used about the same amount? Keep adding tracks one at a time and see if the usage changes between them.

My reasoning here is that as Live keeps one core in use per track that tracks with about the same load should be spread over the cores until you have more tracks than cores. After this additional tracks will share cores. At any given point a single core may be overloaded if it is given too much work at one time.

Therefore a quadcore processor that is faster at executing code in a single core — i e actually being faster and not the processor speed setting — could be faster than a slower per core processor that has additional cores to share the burden. There's also some overhead with many cores and different types of bottlenecks that may cause wasteful wait states.
Make some music!

theinfinate
Posts: 196
Joined: Sat Jan 22, 2011 10:21 am

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by theinfinate » Fri Aug 26, 2016 2:35 am

Stromkraft wrote:
theinfinate wrote: a Mac Pro tower as it can be customised, I thought it was simple maths behind the processing power being faster as in 4 x 2.6 vs 12 x 3.3, I thought that was a no brainer??
The processor speed is not as interesting as how much machine code can be executed in how many threads in a given cycle or specific time frame, which is why bench scores and machine comparisons should be looked up before making a purchase. In theory, 12 real cores should be able to outperform 4 cores even if a little slower, provided one core isn't overloaded and all cores can actually be utilized. There are also generational differences between processors. A 2009 CPU is very old (7 years!) in the computer world.

Can you for instance do this test or similar:
  • Set your instruments in single core mode if they have such a setting.
  • Create as many similar instrument tracks as you have real cores — not counting virtual — playing a longer section, like 32 bars, with 16ths. Make the tracks somewhat different to each other in the MIDI events.
  • Set the buffer size so that playback is smooth.
  • Open up Activity Monitor and press Cmd-2 and look over how the cores are used.

Are the cores used about the same amount? Keep adding tracks one at a time and see if the usage changes between them.

My reasoning here is that as Live keeps one core in use per track that tracks with about the same load should be spread over the cores until you have more tracks than cores. After this additional tracks will share cores. At any given point a single core may be overloaded if it is given too much work at one time.

Therefore a quadcore processor that is faster at executing code in a single core — i e actually being faster and not the processor speed setting — could be faster than a slower per core processor that has additional cores to share the burden. There's also some overhead with many cores and different types of bottlenecks that may cause wasteful wait states.
Thankyou so much for your detailed response, just to make it worse, the guy I bought it off forgot to put a wifi card inside so I'm having trouble installing all my software, loosing patience with this guy! Still waiting for the wifi card to be sent to me so il do these tests and a geek bench test when I get it up and running

Quez
Posts: 142
Joined: Sun Jan 06, 2013 5:53 pm
Contact:

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by Quez » Sun Aug 28, 2016 4:20 pm

The Mac Pro has not been updated in a long while. I wouldn't buy one right now.

Machinesworking
Posts: 11118
Joined: Wed Jun 23, 2004 9:30 pm
Location: Seattle

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by Machinesworking » Mon Aug 29, 2016 4:56 pm

theinfinate wrote:I've been saving for a while now to buy a mac pro as my macbook pro (i7 2.6 quad) suffers recording at 96 kHz and having multiple instances of omnisphere, contact, addictive drums etc. I just received my mac pro 2009 2x 6 core 3.33 ghz thinking it would do the trick. I don't know if i haven't configured something correctly but with 3 instances of omnisphere open on both computers, macbook pro is sitting at 15 percent and the mac pro is sitting at 22 percent. Can someone she some light on this situation?
OK, this is one of the worst myths that gets perpetuated ad infinitum on this site, that a non to failure test indicates anything at all.

Let me explain, your laptop has four cores, the CPU in them is newer, and if you could magically make it so that the Mac Pro and the Macbook Pro were only able to use one core then I'm sure that the Macbook Pro would come out ahead. The Xeon chips are designed for servers, steady power etc. the i7 is designed for fast consumer experiences, it performs better in single core tests as mentioned. Live can use multiple cores, so none of that matters that much. Lives CPU meter will read out like it's maxed out before it seems to use all available power. Here's the rub though, Live will continue to read at roughly 70% at up to 5 times as many plug ins as when it reads at 65%, Ableton seems to have designed the meter in the Rhicter scale, I have no idea why except to say that it for dammed sure is not optimized for multiple CPUs...

The Live Performance Test that gets bounced around this site is a total waste of time, Lives performance CPU wise is seemingly random when not at right below failure. I tested a 2009 mac pro and it gave it an abysmal score, in par with a mac mini, doing the test to failure (running dozens of plug ins until it crackles the audio then backing off until the audio is clean), the mac pro trounces the mini in that test, easily five times as many plug ins could be run. A quad core macbook pro is going to do well, but a 12 core mac pro will very easily beat it in stressful high plug in count situations, that's all there is to it.

A good reference for real world approximations of performance is Geekbench.
https://browser.primatelabs.com/mac-benchmarks
I'm not at my desktop, but I have a similar set up to yours, a 2009 mac pro modded with 12 cores at 3.33ghz and a 2012 macbook pro with a 4 core 2.7 i7.
The Macbook does really well considering, but it's still maybe 2/3 of what the mac pro can handle, especially in high track count projects.

Single core tasks (a single track in Live) will look better for the macbook pro, but once you add in a dozen tracks with plug ins etc. the mac pro will beat the macbook pro. The modern i7 chips are faster in single core use, but a quad 2.6 in no way is going to have the track count of a 12 core 3.33, even if there are architectural improvements in the chips themselves. To really sink this in, the modifications to my machine put it on par with the fastest mac pro garbage can out today, for audio applications, CPU is everything, I'm certain running Final Cut Pro or Premier on the garbage can mac pro is better, and not relevant to what we do. Thinking somehow that faster RAM speed or other improvements are going to beat pure CPU power for processor intensive software like Live is barking up the wrong tree.

Machinesworking
Posts: 11118
Joined: Wed Jun 23, 2004 9:30 pm
Location: Seattle

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by Machinesworking » Mon Aug 29, 2016 5:56 pm

To really sink this in, current geekbench single and multi core scores:

My 3.33ghz 12 core 2009 mac pro
Multi - 30,424
Single 2,571

Compared to a 'comparable to yours' macbook pro.
2013 Retina quad 2.6
Multi 12,358
Single 3,701

In single core performance, not to stress i.e. running three tracks that stress only three cores, you're going to see the single core difference there.
The Live CPU meter is even showing the relative CPU difference if both machines were single core, the i7 is looking roughly 1 1/2 times faster.
It will look like the macbook pro is the superior machine in non stress CPU tests, but in multiple track situations the mac pro will be up to 2 1/2 times more powerful, and this plays out accordingly in the real world.

theinfinate
Posts: 196
Joined: Sat Jan 22, 2011 10:21 am

Re: Macbook pro vs Mac pro, my Macbook pro is faster??

Post by theinfinate » Mon Aug 29, 2016 8:06 pm

Machinesworking wrote:To really sink this in, current geekbench single and multi core scores:

My 3.33ghz 12 core 2009 mac pro
Multi - 30,424
Single 2,571

Compared to a 'comparable to yours' macbook pro.
2013 Retina quad 2.6
Multi 12,358
Single 3,701

In single core performance, not to stress i.e. running three tracks that stress only three cores, you're going to see the single core difference there.
The Live CPU meter is even showing the relative CPU difference if both machines were single core, the i7 is looking roughly 1 1/2 times faster.
It will look like the macbook pro is the superior machine in non stress CPU tests, but in multiple track situations the mac pro will be up to 2 1/2 times more powerful, and this plays out accordingly in the real world.
Thankyou so much for your info, makes me feel better! I was hoping this might be the case. Still waiting on a wifi card to activate a few things, really want to test this bad boy out!

Post Reply