Mac Mini 2.6 i7 Quad = Mac Pro 3.33 Xeon Hex

Macintosh software/hardware discussion and troubleshooting

Moderator: James Steele

Post Reply
Trevor_NY
Posts: 5
Joined: Thu Jan 25, 2018 3:53 am
Primary DAW OS: MacOS

Mac Mini 2.6 i7 Quad = Mac Pro 3.33 Xeon Hex

Post by Trevor_NY »

Hi all,

Thought this might be of interest.

I created a DP project using lots of MOTU only synths and plugins, as a stress test. Lots of FX on all synth channels too..

Loaded project up on two machines (both running High Sierra) and with fresh 256GB SSDs in each (no other drives - i.e project was running off OS drive in both)

Specs as follows:

MacMini 2012 2.6ghz i7 Quadcore 16GB (1600mhz) ram 256GB SSD

MacPro 2010 3.33ghz Xeon Hexcore 32GB (1333mhz) ram 256GB SSD

Ran tests at 64 buffer, 24bit 44.1khz.

CPU meter on both machines was playing between 50% and 75%.

Playback buffer on MacMini around 25%, on MacPro around 35%.

I am shocked! I really expected the hex core to trounce the MacMini.

Next step was to lower the buffer to 32..I thought the grunt of the MacPro might see it beat the Mini here - no - same result. Both CPU meters between 90% and pegged at 100% RED.

What gives? Any thoughts on this?

So, do I have to upgrade my workstation class 5,1 Mac Pro to a 12 core in order to beat a lowly 4 core Mac Mini with laptop innards?!

NOTE: No audio tracks playing - this was simply a realtime test of synths/plugins.
User avatar
HCMarkus
Posts: 9713
Joined: Tue Jan 10, 2006 9:01 am
Primary DAW OS: MacOS
Location: Rancho Bohemia, California
Contact:

Re: Mac Mini 2.6 i7 Quad = Mac Pro 3.33 Xeon Hex

Post by HCMarkus »

Very interesting! It seems unlikely that one generation newer CPU would make up the difference in both clock speed and core count, but it is hard to argue with objective test results.

What about with larger buffers? I personally run a buffer no smaller than 128, and bump up to 1028 when mixing.

I wonder if a previous version of OS might work better on these older machines.

It may also be instructive to take a look at Apple's Activity Monitor's reported CPU usage

I assume you are running DP9.51

As for your question, the answer depends on whether you need more oomph than you are currently getting from your hex MP to run the projects you want to run.
HC Markus
M1 Mac Studio Ultra • 64GB RAM • 828es • macOS 13.6.4 • DP 11.31
User avatar
mikehalloran
Posts: 15134
Joined: Sun Jan 25, 2009 5:08 pm
Primary DAW OS: MacOS
Location: Sillie Con Valley

Re: Mac Mini 2.6 i7 Quad = Mac Pro 3.33 Xeon Hex

Post by mikehalloran »

You’re not taxing either machine. Both are fast. The Mini has a faster bus speed and integrated GPU.

Run 6 instances of Kontakt or edit video and see what happens.

My 2010 iMac on High Sierra with SSDs runs faster and cooler on the same machine performing the same task on El Cap. That 10.13 boots from an 850 EVO and 10.11 boots from an MX300 is insignificant as far as I can tell. APFS does make a difference.

The only reason I keep El Capitán on one drive is for testing—some of my school clients still have 24” iMacs and can’t upgrade the OS (the hacks on the internet don’t work on these machines). Not going backwards.
DP 11.31; 828mkII FW, micro lite, M4, MTP/AV USB Firmware 2.0.1
2023 Mac Studio M2 8TB, 192GB RAM, OS Sonoma 14.4, USB4 8TB external, M-Audio AIR 192|14, Mackie ProFxv3 6/10/12; 2012 MBPs Catalina, Mojave
IK-NI-Izotope-PSP-Garritan-Antares, LogicPro X, Finale 27.4, Dorico 5.2, Notion 6, Overture 5, TwistedWave, DSP-Q 5, SmartScore64 Pro, Toast 20 Pro
User avatar
Michael Canavan
Posts: 3540
Joined: Fri Jul 15, 2005 10:01 pm
Primary DAW OS: MacOS
Location: seattle

Re: Mac Mini 2.6 i7 Quad = Mac Pro 3.33 Xeon Hex

Post by Michael Canavan »

The only true test of a computers possible performance via a test is a test to failure.
This involves loading plug ins on dozens of tracks playing audio and or MIDI files. The key is to load and distribute until audio has noticeable artifacts then pull plug ins until the DAW is stable.

Last time I did a comparative DAW 101 style test with DP7 and Logic 8, Logic looked like a clear winner up until about 60% on the cpu meters. At that point DP stayed around 70% for nearly twice as many many plug ins before crackling artifacts showed up in the monitors. Basically the performance meter of a DAW or a brand of cpu in this case is not a linear path, so tests not to failure don’t reveal anything.

merry It’s not just a DP or DAW thing, in not to failure tests a mac mini used “less” cpu in Ableton Live than my juiced up mac pro, but in tests to failure the mac pro easily handles 3-5 times as many plug ins.
M2 Studio Ultra, RME Babyface FS, Slate Raven Mti2, NI SL88 MKII, Linnstrument, MPC Live II, Launchpad MK3. Hundreds of plug ins.
Trevor_NY
Posts: 5
Joined: Thu Jan 25, 2018 3:53 am
Primary DAW OS: MacOS

Re: Mac Mini 2.6 i7 Quad = Mac Pro 3.33 Xeon Hex

Post by Trevor_NY »

I have taken 15 seconds of video showing both the Mini and Pro CPU meters and Activity Monitor.

Both at 24bit, 44.1k and 64 sample buffer, using Built-In output.



https://www.youtube.com/watch?v=_vQxMTKLUrE

https://www.youtube.com/watch?v=wh99GwZyodE
User avatar
zaratero
Posts: 453
Joined: Tue Aug 05, 2008 4:50 pm
Primary DAW OS: MacOS
Location: Gasteiz

Re: Mac Mini 2.6 i7 Quad = Mac Pro 3.33 Xeon Hex

Post by zaratero »

Hi, after reading this post, I did something alike, only I simply took the last documentary I did post on and added pluggins in a sort of failure test as well.

MacPro 3.1 octo 2.8 - 16 GB - SSD drive for OS & HDD for Audio & Video
MacMini 6.1 dual i5 2.5 - 16GB - SSD for OS & SSD for Audio & Video
single 2k screen for this test on both

Roughly 30 tracks with Comp/Exp & EQ on each channel plus some verbs. Denoisers, compressors & limiters in the auxs.... 90 minute project with a ton of automation. Around 10GB for 2000 audio files. 1080p H264 video. Tested at both 512 and 1024 buffer on internal soundcards.

The MacPro will take an overall heavier load
The Mini will take a heavier load in a single aux channel
File open time is 5 minutes in MacPro vs 4 minutes in Mini
Video is smoother in the Mini, seems to hold on tighter until the whole thing is overloaded. MacPro wobbles a little earlier when scrolling and using the shuttle.
Both show the same zooming/scrolling sluggishness if I try to display too much content in the SEQ with a slight advantage for the MacPro here. (512Mb card vs no GPU). Both will stop being responsive with big amounts of data on screen before they crackle, so while I´m using them both feel to me more limited by the graphic performance of DP rather than the audio processing capability itself.
I believe the MP has an edge regarding power after all, but they surprisingly perform very similar, to my needs this is.

Now, this thing runs hot as hell, although after playing in a loop for 2 days at that load and those temps, I see no performance degradation when I reengage work. I wonder about lifespan though.
I was sort of eskeptical, but glad it works this good since I put it together for 480€ with brand new RAM and SSDs. It makes for a pretty decent fully silent computer, which is what I was after.

This post gave me some motivation to engage in this Mini build, so thanks.
http://www.cueaudio.org
Macpro 3.1 Octocore-16Gb RAM-OSX10.8 - DP 7.24
Macbook pro Core2Duo-4Gb RAM-OSX10.6 - DP 7.24
User avatar
Michael Canavan
Posts: 3540
Joined: Fri Jul 15, 2005 10:01 pm
Primary DAW OS: MacOS
Location: seattle

Re: Mac Mini 2.6 i7 Quad = Mac Pro 3.33 Xeon Hex

Post by Michael Canavan »

zaratero wrote: MacPro 3.1 octo 2.8
MacMini 6.1 dual i5 2.5
...........
The MacPro will take an overall heavier load
The Mini will take a heavier load in a single aux channel
This makes perfect sense, the mini is a newer chip by at least a couple years, and the pro has twice the cores. So even though it's rated at slower speeds, other factors in the chip matter and contribute to it's better performance on the aux channel. In a single track, chip speed counts, over multiple tracks, multiple cores count.
M2 Studio Ultra, RME Babyface FS, Slate Raven Mti2, NI SL88 MKII, Linnstrument, MPC Live II, Launchpad MK3. Hundreds of plug ins.
Trevor_NY
Posts: 5
Joined: Thu Jan 25, 2018 3:53 am
Primary DAW OS: MacOS

Re: Mac Mini 2.6 i7 Quad = Mac Pro 3.33 Xeon Hex

Post by Trevor_NY »

UPDATE:

I have just bough an entry level i5 iMac 27" 3.4ghz with 8GB ram.

Gobsmacked when I tried to run the same project as per MacMini 2.6 quad i7 with 16GB RAM and MacPro 3.33 Hex with 32GB RAM.

THE IMAC TROUNCES THEM BOTH!

I was running at 32 sample buffer and it was cruising through at 50% CPU load and 20% playback buffer. The other two could barely play the project at 32 samples ..stuttering, crackling and with CPU bouncing between 90% and 100% pegged red.

So, a piddly non hyper threaded i5 with a measly 8 GB ram and Fusion drive basically eats the six-core Xeon MacPro (running the project off a PCI mounted SSD mind you) alive..what is the world coming to?
User avatar
HCMarkus
Posts: 9713
Joined: Tue Jan 10, 2006 9:01 am
Primary DAW OS: MacOS
Location: Rancho Bohemia, California
Contact:

Re: Mac Mini 2.6 i7 Quad = Mac Pro 3.33 Xeon Hex

Post by HCMarkus »

MY world is not running at a buffer of 32!
Post Reply