The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Logic Pro Multicore Benchmarktest ! Audio Interfaces
Old 18th January 2018
  #3151
Quote:
Originally Posted by chevaliermusic View Post
Hi - Here are the (88.2 sample rate) results for my 2013 Mac Pro 12 core:

184 tracks using the following settings:
Threads: 24 (Automatic)
I/O Buffer: 64
Process Buffer: Medium

Audio Interface: Antelope Audio DISCRETE 8 88.2 khz sample rate setting

Does that mean 368 tracks at 44.1? (theoretically)
At least on my iMac Pro it didn’t translate to a 1:1 relationship with the sample rate to track count. I would guess you get about 60-70% more tracks at 44.1.
Old 2nd April 2018
  #3152
Lives for gear
I9 7980xe hack (imac pro smbios) pro this thread-

How to build your own iMac Pro [Successful Build/Extended Guide] | tonymacx86.com

It plays 255 tracks at 96k (and shows about 40-50% load) so I bumped it up to 192k and got 180 tracks!

I'm not sure what that translates to at 44.1, but it seems pretty comfortable in the 600-700's. At some point this test isn't very good if it's only maxing out a specific bottleneck like cache etc.

Also looking at the intel power gadget, LPX (or any other daw) is not doing a very good job of loading the processor. Utilization is only around 40%, CPU temps in the mid 40's and power use in the 150's. Under a stress test I'm putting a 4.4ghz cap (all cores) and stable in the mid 80s and whopping 280 watts!

This machine ROCKS! Thunderbolt working very smoothly with a Red 8Pre as well as a dock with 4k video, usb, NIC. From my early test I can actually bridge one of the built in NICs on the mobo with the one on the dock and the TB cable will carry Dante too! That means I can have a single (fiber optic!) connection from machine room to control rooms.

Also the Aquantia 10G NICs are working with a TP link 10g hub that supports QOS and looks like dante is all running clean there so far. More testing to do, but for my needs it's like all this gear fell out of the sky right at the time I needed it.
Old 2nd April 2018
  #3153
Lives for gear
Heres GB3 /32

This thing is a monster
Attached Thumbnails
Logic Pro Multicore Benchmarktest !-screen-shot-2018-04-02-1.09.23-pm.png  
Old 3rd April 2018
  #3154
Gear Nut
 

Hackintosh i7 8700k 6 cores 12 threads at stock clock
High Sierra 10.13.3
Audio interface MOTU 828 MK2

Logic Pro X 10.4.0

Threads: 12 (Automatic)
I/O Buffer: 64
Process Buffer: Medium
88.2 sample rate

110 tracks
_________________________

44.1 sample rate
I/O Buffer: 256

200 tracks
__________________________

Quote:
Originally Posted by RyanC View Post
Also looking at the intel power gadget, LPX (or any other daw) is not doing a very good job of loading the processor. Utilization is only around 40%, CPU temps in the mid 40's and power use in the 150's
With 200 tracks Intel power gadget tell me that the cpu utilization it’s at 75%, temps around 50°/54°
Old 3rd April 2018
  #3155
Lives for gear
I worked with some of the settings in the BIOS (mainly the AVX offsets), now I'm able to get a stable 200 tracks at 192khz. I'm still capping it at 4.4Ghz, I think this machine would run fine higher, but I prefer to err on the safe side.

Cpu utilization is between 50-60% (I think it might have fluctuated a bit higher than 40% before). Temps are nice and stable around 60C with a Noctua NH-D15.

Quote:
Originally Posted by johannes2510 View Post
With 200 tracks Intel power gadget tell me that the cpu utilization it’s at 75%, temps around 50°/54°
I built a 8700k system first (Asus Prime z370) thinking it could replace my 5960x, there really is no reason not to bump up that clock speed. Mine was stable at 5Ghz with the NH-D15.

The thing about these newer gen processors is that they don't just park at full loads and lock at high clocks all the time...I hope that when apple makes the next mac pro they take this into account. These chips are fully designed to run at higher than stock speeds provided good thermal dissipation. You can see in the LPX test that it is not exceeding it's rated TDP.
Attached Thumbnails
Logic Pro Multicore Benchmarktest !-screen-shot-2018-04-03-7.37.48-am.jpg   Logic Pro Multicore Benchmarktest !-screen-shot-2018-04-03-7.37.42-am.jpg  
Old 3rd April 2018
  #3156
Gear Nut
 

Quote:
Originally Posted by RyanC View Post
I built a 8700k system first (Asus Prime z370) thinking it could replace my 5960x, there really is no reason not to bump up that clock speed. Mine was stable at 5Ghz with the NH-D15.

The thing about these newer gen processors is that they don't just park at full loads and lock at high clocks all the time...I hope that when apple makes the next mac pro they take this into account. These chips are fully designed to run at higher than stock speeds provided good thermal dissipation. You can see in the LPX test that it is not exceeding it's rated TDP.
Agreed, but it’s my first hackintosh and never do an overclock, so I am in the study phase

I have a Noctua nh-u14s and I suppose that when I will take courage and try overclock. I think that I can go a little less than a nh-D15… But I I will reach a 4.8/4.9 I will be happy.

Do you know a reliable guide for overclock an hackintosh with the 7800k with Gigabyte aourus gaming 7 ?
Old 3rd April 2018
  #3157
Lives for gear
Quote:
Originally Posted by johannes2510 View Post
Do you know a reliable guide for overclock an hackintosh with the 7800k with Gigabyte aourus gaming 7 ?
No, but honestly all you have to do is set the voltage to 1.3 or so, set the clock to maybe 4.6 and then monitor temps and adjust the clock from there. I set mine to 1.3, set it to 5ghz and it was good to go...personally I think this logic stress test is a good one to push temps and test for stability if you aren't doing any rendering or other jobs that can reach 100% cpu utilization.
Old 3rd April 2018
  #3158
Gear Nut
 

Ok thanks.
Old 6th April 2018
  #3159
Gear Nut
 

Selfquote:

Quote:
Originally Posted by johannes2510 View Post
Hackintosh i7 8700k 6 cores 12 threads at stock clock
High Sierra 10.13.3
Audio interface MOTU 828 MK2

Logic Pro X 10.4.0

44.1 sample rate
I/O Buffer: 256

200 tracks
__________
I have discovered that my memory 2400 mhz was running at 2133 because XMP was not enabled in BIOS.

After enable XMP, so the memory now run at 2400, Logic has add 14 tracks so now is at 214.

And with the other settings:

I/O Buffer: 64
88.2 sample rate

Logic add 7 tracks so 117 (half of 44.1 setup that it's.... Logic )
Old 7th April 2018
  #3160
Lives for gear
 

Hackintosh i7 6800K 6 cores 12 threads @ 4GHZ (Logic doesn't seem to be maxing out the cores though, they're like 3.6GHZ or so on Intel Power Gadget

Sierra 10.12.6

Audio interface FIREFACE 400

Logic Pro X 10.4.1

Threads: 12 (Automatic)

44.1 sample rate
I/O Buffer: 256

162 tracks

Old 7th April 2018
  #3161
Gear Nut
 

First and quick test overclock 8700K at 4.8 ghz (with the automatic overclock of MOBO Gigabyte).

I/O Buffer 64 / 88.2 sample rate Logic play 129 tracks.
Old 28th April 2018
  #3162
Gear Nut
 

Quote:
Originally Posted by RyanC View Post
No, but honestly all you have to do is set the voltage to 1.3 or so, set the clock to maybe 4.6 and then monitor temps and adjust the clock from there. I set mine to 1.3, set it to 5ghz and it was good to go...personally I think this logic stress test is a good one to push temps and test for stability if you aren't doing any rendering or other jobs that can reach 100% cpu utilization.
I have find a guide for overclock 8700K on Gygabite aorus gaming 7
GIGABYTE Z370 Guide to Overclocking Coffee Lake CPUs to 5GHz+ - Overclocking.Guide
but after that I have find best, a guy (wingman99) at overclock.net that helped me a lot for understand and set a dynamic voltage for overclock that is more safe because don't generate voltage overshoots that is bad for chips.
In short fixed voltage is the old way to overclock, dynamic is more modern and more appropriate for modern chips with all the power saving features etc.
If you are intrested you can go here, starting last post on the page (it's long):

Gigabyte Aorus Gaming 7 / 8700K Overclocking settings - Page 31 - Overclock.net - An Overclocking Community

SO I have made more tests with presets for overclock at 4.7, 4.8 and 4.9.
5.0 it’s a little hot for me so I don’t.

Threads: 12 (Automatic)
I/O Buffer: 64
Process Buffer: Medium
88.2 sample rate
(at CPU stock was 110 tracks.)

4.7 ghz = 127 tracks
4.8 ghz = 129
4.9 ghz = 130

4.9 ghz 48 sample rate = 219 tracks
4.9 ghz 44.1 sample rate = 245 tracks

In the cpu meter the first core (thread?) on the right is always lower than the others. Someone know why?

Old 29th April 2018
  #3163
Lives for gear
Quote:
Originally Posted by johannes2510 View Post
I have find a guide for overclock 8700K on Gygabite aorus gaming 7
GIGABYTE Z370 Guide to Overclocking Coffee Lake CPUs to 5GHz+ - Overclocking.Guide
but after that I have find best, a guy (wingman99) at overclock.net that helped me a lot for understand and set a dynamic voltage for overclock that is more safe because don't generate voltage overshoots that is bad for chips.
Good to know, I actually sold the 8700k system to a friend, built a server with a 8100 (which is super snappy) and then am focusing on the 7980xe systems for the audio workstations. X299 is dynamic voltage only...

But I will check that out and see if I can update the 8700k system my friend has. Gigabyte and Asus are a bit different for OC as well, did the intel power gadget show the processor dropping down to just a few watts at idle?

I'm not sure if I was already running dynamic voltage, but if the volts were fixed at 1.3 then certainly the current would go way down when not under load.
Old 29th April 2018
  #3164
Lives for gear
Self Quote:

Quote:
Originally Posted by RyanC View Post
I hope that when apple makes the next mac pro they take this into account. These chips are fully designed to run at higher than stock speeds provided good thermal dissipation. You can see in the LPX test that it is not exceeding it's rated TDP.
Working with the 7980xe system, what would be great is if Apple, and windows system designers moving forward could start to abandon the idea of a chip having a *clock speed* and evolve the thinking more into the realm of a thermal stability limit.

With low latency processes like audio, we will NEVER reach 100% cpu utilization. So running these systems as if they were render farms is stupid. The clock speed should be pushed up to the thermal stability threshold, as this makes the most overhead for our use.

The traditional X.x Ghz processor model needs to go away.
Old 29th April 2018
  #3165
Gear Nut
 

Quote:
Originally Posted by RyanC View Post
But I will check that out and see if I can update the 8700k system my friend has. Gigabyte and Asus are a bit different for OC as well,
For what I have understand the differences are only the way they name settings but the substance are the same. I think research "8700k asus dynamic-adaptive voltage" you can find the solution.
For example (if I am not wrong and remember that I am a newbe on overclock so don't do that without confirmation) in Gygabite mobos for dynamic IA AC/DC for 8700K must set at 1. For Asus I think is 0.01 but the function is the same.

Quote:
did the intel power gadget show the processor dropping down to just a few watts at idle?
The guy who helped me says that with dynamic method the voltage at idle is ~0.789v

This is my 8700K overclocked at 4.8 ghz with dynamic voltage and CPU at idle dropped to the minimum 0.80ghz (I have see also pkg at 2.60):



Quote:
I'm not sure if I was already running dynamic voltage, but if the volts were fixed at 1.3 then certainly the current would go way down when not under load.
I think it depends if you have disable or not c-states and Intel power features that at default are all enabled (I am not sure about the names of all them). I think that if you disable them, like the guide that I have follow at beginning for fixed mode suggest for some of them, not sure if or in what mesure that will happen.
Old 1st May 2018
  #3166
Gear Addict
 

I see that Apple has a LPX max track count comparison between the 4.2Ghz iMac, 12-Core Mac Pro, 10-core iMac Pro and the 18-core iMac Pro using Logic’s Sculpture synth on the online Apple store iMac Pro overview. The iMac, which is being used as the base, has 8GB of memory and the others have 64 to 128GB.

So it’s 1.8x for the 12-Core Mac Pro, 8.3x iMac Pro 10-Core and 12.4x iMac Pro 18-Core vs the iMac 4.2Ghz.
Old 2nd May 2018
  #3167
Lives for gear
Quote:
Originally Posted by fastlanephil View Post
I see that Apple has a LPX max track count comparison between the 4.2Ghz iMac, 12-Core Mac Pro, 10-core iMac Pro and the 18-core iMac Pro using Logic’s Sculpture synth on the online Apple store iMac Pro overview. The iMac, which is being used as the base, has 8GB of memory and the others have 64 to 128GB.

So it’s 1.8x for the 12-Core Mac Pro, 8.3x iMac Pro 10-Core and 12.4x iMac Pro 18-Core vs the iMac 4.2Ghz.
Do they have a link to download their test file anywhere? Could be a new benchmark for this...
Old 2nd May 2018
  #3168
Lives for gear
 

Can someone explain the extreme performance difference in the Logic Pro X benchmark?

Specs:

iMac Pro 10-core - 2017 iMac Pro 3.0GHz 10-Core Xeon W-2150B CPU, 64GB of 2666MHz DDR4 ECC SDRAM, AMD Radeon Pro Vega 64 (16GB) GPU

iMac Pro 8-core - 2017 iMac Pro 3.2GHz 8-Core Xeon W-2140B CPU, 32GB of 2666MHz DDR4 ECC SDRAM, AMD Radeon Pro Vega 64 (16GB) GPU

iMac Pro 8-core versus 10-core

251 tracks vs 508 tracks ? How is it possible that one can run the double amount of tracks with only 2 more cores??
Old 2nd May 2018
  #3169
Gear Addict
 

Quote:
Originally Posted by RyanC View Post
Do they have a link to download their test file anywhere? Could be a new benchmark for this...
As far as I know, Apple doesn’t provide the testing details or any links with their product comparisons.

I think it does show that the iMac Pro 10-Core “might” be the best bang for the buck but I would think that Apple does more than one type of test and then uses the one that looks best for marketing their product.
Old 6th May 2018
  #3170
Here for the gear
 

I have an i3 Kaby Lake with 8 GB that stopped working at 63. I'll never be close to that many tracks ever lol
Old 13th May 2018
  #3171
Gear Head
 
Konan's Avatar
 

I'm unsure if I am using this right, but my results are-

84 actual tracks load, of these 16 are green, but when I play I get a system overload.

I have to press the button that looks like an on switch off, and mute the tracks, I had to go down to just FIVE tracks so it would play without a system overload.

This is using a 2010 MBP 13" core 2 duo 2.4GHZ 4GB ram.

I am looking at used macs to upgrade, I am going for a 2012 non retina or 2013 retina model with an i7 depending on what deals I can find, I'm not bothered about the display but the ability to upgrade the 2012 non retina is nice.

Will a 16GB mac with a i7 run a lot more tracks than my core 2 duo with 4GB ram?

Will I notice the i7 performs much better for me? I can currently make actual music with around 16 tracks before I hit overload. This is with effects and what I need. Will I easily get 32 or more tracks with the i7?
Old 17th May 2018
  #3172
Gear Head
 
Konan's Avatar
 

Wow, I got my new mac today- it run the whole benchmark of 84 tracks!

My old 2010 MBP could only run 5, and as this ones benchmarks was about 5 times as good, I expected 25, but it ran 84 fine. No way will I need to run anywhere near this amount of tracks with so many effects, I was quite happily making music on my old mac with just 16 tracks, so having this much extra available power is great.

It is a 2013 15" MBP retina, 2.4GHZ quad core i7, 8GB ram and 256BG SSD. It cost me less than half the price of a new base model 13", so seemed like it would be silly to miss the offer.

EDIT: I haven't tried more as it was just 84 that loaded up(EDIT- 99, I didn't try before until I got CPU overload), the CPU looks like it is pretty much at its max in logic with the 8 cores (4, but hyperthreading).

This is with a 128 buffer, 44.1. Very happy, even though I am seeing people getting 200+ with some amazing setups.
Old 17th May 2018
  #3173
Gear Nut
 

Quote:
Originally Posted by Konan View Post
Very happy, even though I am seeing people getting 200+ with some amazing setups.
That's normal. You have a laptop, mine (for example) is a tower with really powerfull and modern cpu/ssd etc.

The important is have what we need
Old 17th May 2018
  #3174
Gear Head
 
Konan's Avatar
 

Quote:
Originally Posted by johannes2510 View Post
That's normal. You have a laptop, mine (for example) is a tower with really powerfull and modern cpu/ssd etc.

The important is have what we need
I hadn't actually tried until system overload, just did and it runs 99 OK, 100 is the magic number though as this gives a system overload.

If I had the space, I would love a tower, but in reality being able to run 99 instruments with multiple effects is pretty insane, even for a laptop.

Definitely what I need, and more

EDIT: How on earth am I getting the same as 2016 and 2017 MBP's? Mine is a 2013 2.4 i7 with 8gb ram!Also this is ris without my interface, would that make a difference or is it really my processor that matters in this test?
Old 7th July 2018
  #3175
Lives for gear
 
Robert Randolph's Avatar
 

So, what do we do if we run out of instrument tracks? I'm up to 256 and still have room left.
Old 7th July 2018
  #3176
Gear Addict
 

Quote:
Originally Posted by Robert Randolph View Post
So, what do we do if we run out of instrument tracks? I'm up to 256 and still have room left.
You could load another app or more until you’re struggling to reach 256 tracks. Then use Apple’s Activity Monitor to see how much of the CPU is being used by the other app(s). Then doing some basic math you might have a rough idea how many tracks it’s capable of reaching without the 256 track limit. Also check memory usage if that could be a factor.
Old 8th July 2018
  #3177
Lives for gear
Quote:
Originally Posted by Robert Randolph View Post
So, what do we do if we run out of instrument tracks? I'm up to 256 and still have room left.
Mostly we've been upping the sample rate. But a 7980xe hack can actually do 255 racks at 192k so we're going to need a new test at some point.
Old 8th July 2018
  #3178
Gear Nut
Sadly the processor gains like you saw going from core2duo to i7 haven’t happened over the last 4-5 years. AMD putting pressure with their new chips should help, but Intel has not had much success with smaller manufacturing, which adds to the lack of improvements. I have the 2013 model and my wife has the 2017. Aside from the Thundebolt 3, SSD speeds and touchbar not much difference.

Quote:
Originally Posted by Konan View Post
I hadn't actually tried until system overload, just did and it runs 99 OK, 100 is the magic number though as this gives a system overload.

If I had the space, I would love a tower, but in reality being able to run 99 instruments with multiple effects is pretty insane, even for a laptop.

Definitely what I need, and more

EDIT: How on earth am I getting the same as 2016 and 2017 MBP's? Mine is a 2013 2.4 i7 with 8gb ram!Also this is ris without my interface, would that make a difference or is it really my processor that matters in this test?
Old 12th July 2018
  #3179
Lives for gear
 
nativeaudio's Avatar
 

Quote:
Originally Posted by RyanC View Post
Mostly we've been upping the sample rate. But a 7980xe hack can actually do 255 racks at 192k so we're going to need a new test at some point.
I guess that would be easy - simply by using the current test and eg. doubling the number of notes (occurring at the same time) in each region.
Old 13th July 2018
  #3180
Gear Nut
Until today. Finally more cores and a big price tag.

Quote:
Originally Posted by superscan View Post
Sadly the processor gains like you saw going from core2duo to i7 haven’t happened over the last 4-5 years. AMD putting pressure with their new chips should help, but Intel has not had much success with smaller manufacturing, which adds to the lack of improvements. I have the 2013 model and my wife has the 2017. Aside from the Thundebolt 3, SSD speeds and touchbar not much difference.
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
Odditory / Music Computers
18
Dr Gruv / Gearslutz Secondhand Gear Classifieds
0
Dr Gruv / Gearslutz Secondhand Gear Classifieds
0

Forum Jump
Forum Jump