LIVE FIVE PERFORMANCE TEST - results here

Erubus · Post by **Erubus** » Wed Sep 28, 2005 11:13 pm

To answer the dual processor question, at least as far as PC's go...I believe Macs work the same but, dont know "For Sure":

Dual processors (each) run a tiny bit slower than single core processors...

I.E. - Top end Intel single is about 3.8 GHz out of the box, while top end
Dual cores are (correct me if I'm wrong) about 2.6 GHz each. So, while a 2.8 GHz totasl sounds impressive versus a 3.8, it's not, here's why:

Programs are sent to processors in "threads" Most programs that aren't built for dual processors, make each thread only as short as it needs to be to carry out that function. So, "threads" are sent "long, short, long, short, long long long short short short" and so on. If the processors are spliting the load, if one keeps getting the long ones randomly, it runs actually SLOWER than a single core, because most of the work is done by a single, slower processor.

Dual core optimized programs shrink the size of threads, even out the spacing, and even the load between the cores. Operating systems that come stock with dual processors, usually are optimized so, are going to show dual core usage. But, that doesnt mean the programs installed are going to follow suit.

Hope that wasnt too much readin for ya and, like I said, it may not apply to Mac's...I havent been to interested in buying one so, haven't looked into it very much. Makes sense that it would work the same though.

montrealbreaks · Post by **montrealbreaks** » Wed Sep 28, 2005 11:37 pm

1) Ableton Live CPU Meter % 32-34
2) Compaq Presario 3240 CA
3) XP Home SP1
4) AMD64 3000+
5) 1.25 GB (1 GB chip, 256 MB onboard)
6) Echo Indigo IO and Onboard Audio with DX made no difference at 512 samples 44.1
7) 7200 RPM, 8meg cache

Dr. Zoiberg · Post by **Dr. Zoiberg** » Thu Sep 29, 2005 2:24 am

I think that the test should be heavier for G5 Macs.

I noticed that in Logic the first plug-in will raise the cpu meter a lot, then every other plug-in will only add a small bit.

This test is good for G4 (yes, they are that slow), but it's not good enough to compare last generation Macs and pcs.

AdamJay · Post by **AdamJay** » Thu Sep 29, 2005 3:20 am

Dr. Zoiberg wrote:I think that the test should be heavier for G5 Macs.

I noticed that in Logic the first plug-in will raise the cpu meter a lot, then every other plug-in will only add a small bit.

This test is good for G4 (yes, they are that slow), but it's not good enough to compare last generation Macs and pcs.

Zolberg,
its a baseline test. and considering many G5's score in the 50% range, i'd say its very fair and heavy enough for G5's

But honestly, i cannot possibly see how this test is "not good enough" for G5s.

It tests various effects and instruments that are all built into Live, and mimmicks a typical user's heavy use of the program as a standalone solution. Now, if the typical user sees that his/her machine gets 40%, and the machines he/she is considering buying get 50% and 25%, then this test gives the perspective buyer performance ratings that he/she can compare and scrutinize. And that is exactly the purpose of the test.

This test has nothing at all to do with Logic.

With regards to some devices using more cpu per plugin when only one is used, vs. when you use multiples. I am aware of this, and that is exactly the reason this test contains 6 instances of 'EQ Four'.

Furthermore, many of the folks at Ableton AG are pleased with the parameters of this test as well as the version 4 test.

Dr. Zoiberg · Post by **Dr. Zoiberg** » Thu Sep 29, 2005 9:58 am

AdamJay wrote:Zolberg,
its a baseline test. and considering many G5's score in the 50% range, i'd say its very fair and heavy enough for G5's

But honestly, i cannot possibly see how this test is "not good enough" for G5s.
It tests various effects and instruments that are all built into Live, and mimmicks a typical user's heavy use of the program as a standalone solution. Now, if the typical user sees that his/her machine gets 40%, and the machines he/she is considering buying get 50% and 25%, then this test gives the perspective buyer performance ratings that he/she can compare and scrutinize. And that is exactly the purpose of the test.

This test has nothing at all to do with Logic.

With regards to some devices using more cpu per plugin when only one is used, vs. when you use multiples. I am aware of this, and that is exactly the reason this test contains 6 instances of 'EQ Four'.

Furthermore, many of the folks at Ableton AG are pleased with the parameters of this test as well as the version 4 test.

Adam, I thought that the G5 and pcs could scale differently with more plug-ins, due to the different architecture, and that the performance could be more similar than it seems.

I didn't mean you had done a bad work, at all.

We could simply take two systems (a mac and a pc) that have scored the same with this test and load them with a BIG project, to see how do they scale and if they get the same score even at around 80% cpu load.

Or we could just make some music instead

:

dkrock3k · Post by **dkrock3k** » Fri Sep 30, 2005 12:41 am

Howdy All

1) Ableton Live CPU Meter % 103, peak
2) Laptop/Desktop? Make/Model iMac G4 (sunflower)
3) Operating System OSX10.3.9
4) CPU Make Model and Speed PowerPC G4/ 1Ghz.
5) Amount of Ram / Speed of Ram (if you know) 1GB /133 Mhz.
6) Soundcard (Stock or add on?, usb/firewire/pci ?) CoreAudio
7) Hard Disk Drive Speed (if you know) 7200

Ouch, man. This thing crushed my poor lil machine. I dare not run it on my Laptop, a puny 900Mhz. Compaq.

Good job. Looks like I need to upgrade to a new(er) Mac

Take Care
DJrock3k

Jesper · Post by **Jesper** » Fri Sep 30, 2005 12:46 pm

Hi all

1) Ableton Live Demo CPU Meter 31-35%
2) Self-assembled desktop
3) XP Pro
4) P4 2.4 OC:d to 3.2
5) 512/~400 Mhz dont remeber
6) EgoSys WaMi Rack 24
7) 7200 8Mb

note: output buffer size 960 samples necessary to avoid glitches

Edit: I disabled hyperthreading in BIOS - now I get peaks at 31%

mike holiday · Post by **mike holiday** » Fri Sep 30, 2005 6:18 pm

1) Ableton Live CPU Meter % 62%--peak
2) Laptop/Desktop? Desktop upgraded quicksilver G4
3) Operating System 10.3.9
4) CPU Make Model and Speed: powerlogix duel 1.8 G4
5) Amount of Ram / Speed of Ram (if you know): 768/133 SDRAM
6) Soundcard (Stock or add on?, usb/firewire/pci ?): Digi 001 pci
7) Hard Disk Drive Speed (if you know): 7200/8m

*note: same system was hitting 126+% with the stock 800 mhz cpu with max buffer size---i'v seen this system hit 200% ouch glad my speakers take more then my amp gives

the duel is nice i can run 85% + spikes without glitches and beachballs!!

JJupiter · Post by **JJupiter** » Sat Oct 01, 2005 9:02 am

1) Ableton Live CPU Meter % 37%
2) Laptop/Desktop? Apple PowerMac
3) Operating System 10.4.2
4) CPU Make Model and Speed G5 2x 2.5GHz
5) Amount of Ram / Speed of Ram? 2.5 GB DDR SDRAM PC3200U-30330
6) Soundcard (Stock or add on?, usb/fire wire/pci ?) M-Audio FW410 FireWire
7) Hard Disk Drive Speed ? 7200/8M

Jubei · Post by **Jubei** » Mon Oct 03, 2005 10:21 pm

Alright Adam, I could use a little help as to why I have the worst numbers running this test, and yet when I look at the specs they are not that off? First things first:

1) Ableton Live CPU Meter - 130% (WOWZERS!!)
2) Laptop/Desktop? Make/Model - Desktop/ imac g4 17inch
3) Operating System - 10.3.9
4) CPU Make Model and Speed - G4 1.25 ghz
5) Amount of Ram / Speed of Ram - 768/DDR333
6) Soundcard - Focusrite Saffire
7) Hard Disk Drive Speed - 7200 rpm

Now, maybe I do not have my machine optimized. I thought I did, but maybe you could help. I look at the specs of what you are using and are we that much different? You are always such a big help.....anyway you can help me get this number down?

camarao · Post by **camarao** » Mon Oct 03, 2005 11:59 pm

1) Ableton Live CPU Meter: 80% !!! (why?!)
2) Make/Model: Apple Powerbook G4 15"
3) Operating System: OSX 10.4.2
4) CPU Make Model and Speed: G4 1.67ghz
5) Amount of Ram / Speed of Ram: 1.5GB / DDR333
6) Soundcard: Stock Core Audio card
7) Hard Disk Drive Speed: 5400rpm

Is it bad ram ? Should I reinstall OSX ? Please help ...

subterFUSE · Post by **subterFUSE** » Tue Oct 04, 2005 1:54 am

Is it bad ram ? Should I reinstall OSX ? Please help ...

Not your RAM... you've got plenty. PowerBooks are just slow. This test is showing why Apple is switching to Intel CPUs.

wavejumper · Post by **wavejumper** » Tue Oct 04, 2005 8:30 am

or it shows that the ableton code is not optimised to run on OSX. despite their claims to the contrary.

Winterpark · Post by **Winterpark** » Tue Oct 04, 2005 8:36 am

wavejumper wrote:or it shows that the ableton code is not optimised to run on OSX. despite their claims to the contrary.

who's claiming?

subterFUSE · Post by **subterFUSE** » Tue Oct 04, 2005 1:24 pm

No. It is just a simple FACT that the G4 processor found in the PowerBooks is no longer fast by today's standards. What's so hard for people to understand about this? It's not some huge revelation.

If you like Mac... wonderful. But let's just be honest. The PowerBooks are very outdated technology. That's the reason they are not scoring well on this test. They just don't have the CPU power of a G5, or PC alternatives.

What is so wrong with saying the obvious?

LIVE FIVE PERFORMANCE TEST - results here

or

Re: or