more cores or more clock speed? for DP

[bayswater]

What did we learn? That SSDs, more memory and buffer size have more to do with performance than the CPU? OK... That's not news, really,

Not so much news, but the point that came out for was that if you have a system level process that ties up the CPU unnecessarily, it doesn't matter if you have SSDs, big buffers, TB, and a zillion gigahertz kilocore CPU, you'll get glitches. Increasing the buffer size gets around problems, but a properly designed and integrated system is the better solution.

[bayswater]

These are a few years old, but will give you an idea of realtime benchmarks:
http://www.dawbench.com/win7-v-osx-1.htm

Page 6 of these tests covers VI performance. I would be interested to see recent numbers with Win10 and OSX. As OSs, cpu families and builds change, performance can as well. To me, the tests of most interest are audio interface comparisons. Performance can vary wildly from very good low latency performance with RME, to poor performance with some low cost USB interfaces.

I remember this when it first appeared. It seems to tell you more about the DAW and how well it works on the OS, than about the underlying real time performance of various systems. E.g. it shows up how, at the time, Cubase on the Mac was pretty much useless.

[toodamnhip]

What did we learn? That SSDs, more memory and buffer size have more to do with performance than the CPU? OK... That's not news, really,

Not so much news, but the point that came out for was that if you have a system level process that ties up the CPU unnecessarily, it doesn't matter if you have SSDs, big buffers, TB, and a zillion gigahertz kilocore CPU, you'll get glitches. Increasing the buffer size gets around problems, but a properly designed and integrated system is the better solution.

+1
Now the question is--HOW to do this!

[toodamnhip]

These are a few years old, but will give you an idea of realtime benchmarks:
http://www.dawbench.com/win7-v-osx-1.htm

Page 6 of these tests covers VI performance. I would be interested to see recent numbers with Win10 and OSX. As OSs, cpu families and builds change, performance can as well. To me, the tests of most interest are audio interface comparisons. Performance can vary wildly from very good low latency performance with RME, to poor performance with some low cost USB interfaces.

I remember this when it first appeared. It seems to tell you more about the DAW and how well it works on the OS, than about the underlying real time performance of various systems. E.g. it shows up how, at the time, Cubase on the Mac was pretty much useless.

What I am trying to figure out here is would it be a wise investment to get a NEWER MOTU audio interface, I have a souped up mid 2010 Mac Pro so I can;t get the new thunderbolt. But the question is, does the new thunderbolt protocol improve on this area called “real time” performance, and how does it compare to an pld 828 mk II. I’d love MOTU to comment on this.
Also, are any of the new motto interfaces better in real time even if used without thunderbolt?

And the last question, if you had to budget between a new MOTU interface of SSD drives, what would you improve1st? I would guess SSD drives are the best bang for the buck available in this search for “perfect” real time performace.

[bayswater]

Now the question is--HOW to do this!

That's why I was asking about benchmarks. Seems to me you need to know the performance of the system as a whole, not just some of the parts.

[bayswater]

What I am trying to figure out here is would it be a wise investment to get a NEWER MOTU audio interface, I have a souped up mid 2010 Mac Pro so I can;t get the new thunderbolt. But the question is, does the new thunderbolt protocol improve on this area called “real time” performance, and how does it compare to an pld 828 mk II. I’d love MOTU to comment on this.
Also, are any of the new motto interfaces better in real time even if used without thunderbolt?

And the last question, if you had to budget between a new MOTU interface of SSD drives, what would you improve1st? I would guess SSD drives are the best bang for the buck available in this search for “perfect” real time performace.

What that video suggests is that we would need to know what exactly the CPU deals with. Does it deal with the hard disk or SSD, or a disk controller? If it's the controller, the SSD might improve the rate of data transfer but not necessarily other aspects of real time performance. On the user interfaces, is the CPU dealing with the MOTU hardware or CoreAudio? If it's CoreAudio, again, real time performance might not improve that much.

I'd want to know more things along these lines before I started buying a lot of stuff expecting it to make a difference.

[toodamnhip]

OK great so is there a way to determine what's eating up real-time processor speed? Do I have to just strip my system down and slowly adding one piece of my system at a time like idiot or is there a more technical wy to see a readout of the real time demands and what those demands are? Stripping ones system down and adding one drive or component at a time is not necessarily a simple proposition. I have tried to watch Activity monitor but haven't quite figured out what exact to look for to see what is tying up the processor.

[bayswater]

I don't think there will be an easy way to do it, other than asking what systems people have, what the specific components are, and how well the overall system performs. I remember the same discussion yeas ago with Windows PCs and Cubase.

[toodamnhip]

I don't think there will be an easy way to do it, other than asking what systems people have, what the specific components are, and how well the overall system performs. I remember the same discussion yeas ago with Windows PCs and Cubase.

Sounds like the solutions are anecdotal....isn;t this supposed to be a HIGHLY technical situation? Why would we need to ask buddies how their systems are doing and chase down shadows to make our computers run at a high level? Maybe I’ll have to look for an online tutorial on optimizing real time performance...one made by a professional computer scientist who understands audio demands.

[mikehalloran]

These are a few years old, but will give you an idea of realtime benchmarks:
http://www.dawbench.com/win7-v-osx-1.htm

Page 6 of these tests covers VI performance. I would be interested to see recent numbers with Win10 and OSX. As OSs, cpu families and builds change, performance can as well. To me, the tests of most interest are audio interface comparisons. Performance can vary wildly from very good low latency performance with RME, to poor performance with some low cost USB interfaces.

I remember this when it first appeared. It seems to tell you more about the DAW and how well it works on the OS, than about the underlying real time performance of various systems. E.g. it shows up how, at the time, Cubase on the Mac was pretty much useless.

If we get back to the OP, that video is little more than a distraction. Except for reminding us about some of the basics, it has nothing to do with how DP performs on a Mac.

For example, the newest iMac i5 running DP will get its butt kicked by a 2009/10 iMac if it has an i7 and the drive is updated to an SSD. I don't care about the numbers or bench scores here. I just need to run the same project on both -- and I have. Now, the latest 5K iMac i7 with a 1T blade... Yes, that's an upgrade but, unless my machine dies tomorrow and I need something in a hurry, my next Mac will be a Mac Pro. Why? I've been using one for a project in another studio ... it's better and I want one.

[bayswater]

I don't think there will be an easy way to do it, other than asking what systems people have, what the specific components are, and how well the overall system performs. I remember the same discussion yeas ago with Windows PCs and Cubase.

Sounds like the solutions are anecdotal....isn;t this supposed to be a HIGHLY technical situation? Why would we need to ask buddies how their systems are doing and chase down shadows to make our computers run at a high level? Maybe I’ll have to look for an online tutorial on optimizing real time performance...one made by a professional computer scientist who understands audio demands.

Unless someone has done a systematic review, yes, it will be anecdotal. One of the things in the video that bears on this: most applications other than audio don't care about real time performance. For video it doesn't matter as much, and that seems to be where Apple has it eggs. If its only audio where this is critical, there might not be much organized research.

In the PC world there are makers that specialize in DAW machines. Maybe it's worth looking at the components they use, and what makes up a good one. Maybe one or more of them do custom Mac builds for Logic users.

There are technical articles on real time computer performance, but they look at applications where "real time" means things like 50 msec response to alert signals, a long time when you're running with a 128 sample buffer.

[bayswater]

If we get back to the OP, that video is little more than a distraction. Except for reminding us about some of the basics, it has nothing to do with how DP performs on a Mac.

That's a curious comment. The same factors apply to a Mac. Maybe it doesn't matter a lot right now because Macs are well designed, and don't often have any notable weak spots. It did matter a while back when poorly implemented USB first appeared on Macs. And it applies when you're pushing a Mac to it's limits, if there is a component that is competing for CPU attention that could be replaced by a component that needs less care and attention.

For example, the newest iMac i5 running DP will get its butt kicked by a 2009/10 iMac if it has an i7 and the drive is updated to an SSD. I don't care about the numbers or bench scores here. I just need to run the same project on both -- and I have. Now, the latest 5K iMac i7 with a 1T blade... Yes, that's an upgrade but, unless my machine dies tomorrow and I need something in a hurry, my next Mac will be a Mac Pro. Why? I've been using one for a project in another studio ... it's better and I want one.

I agree. I have a couple of i5s and and an i7. No question the i7 is better, at least on DP. I don't see such a big difference with Logic - have you done that comparison? Don't know about the Mac Pro -- we do hear from a few disappointed users, and it doesn't seem like it's all paradise. But to get back to TDP's comment, this is all anecdotal, isn't it.

[HCMarkus]

It would seem that a lot of this comes down to how well multithreaded the OS is… Just like DAWs put more than a singe core to work at a time, the OS should allow individual cores to address demands imposed by system components while simultaneously allowing other cores to address only time-specific demands.

Bigger buffer is the solution to the problem; a small buffer is needed only when recording VI's or monitoring through effects when tracking. If an occasional overload occurs while recording MIDI performances with VIs, playback should be fine after buffer size is upped.

OS and all files used by DP should be on SSDs to minimize the time wasted when changing buffer.

[mikehalloran]

For me, Logic exists so that I can convert projects to DP. Fortunately, it's relatively inexpensive. I have never completed a project in Logic.

Of course it's anecdotal. We're a user group. The best reports are from those using it.

I was hoping to get my SSDs installed on my 2010 and 2011 iMacs this weekend but one of my helpers is in the hospital and my other spent most of the weekend sitting with her -- I spent a lot of time there myself.

My 2010 is running via eSATA so, except for running cooler, I don't expect a change when I install it as the 2010 is limited to SATA II.

My 2011 is another story. I am booting from the SSD in a USB 2 dock and the machine is a little faster than from the internal HD. In addition, I was able to run sudo trim force enable and it shows TRIM enabled in About this Mac. Huh? I didn't see that coming. I expect a nice bump when this is installed and running via SATA III.