Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Audio Decoding KILLS MacPro (2009) green factor

Animalk said:
I have been following this thread on and off since it began.

I'm not sure if it has already been posted in this thread, but you guys have Arstechnica's attention.

http://arstechnica.com/apple/news/2010/02/nehalem-mac-pros-take-20-performance-hit-when-playing-audio.ars
Click to expand...

Kudos to Chris Foresman at Ars Technica for bravely writing about this Nehalem (2009) Mac Pro problem. He embodies what a journalist should be: curious and writing about real stories--not kowtowing to powerful corporate interests (like Apple).

I would recommend that everyone here link to the story on as many websites as possible so that its profile is raised and that it eventually makes it to the google news's of the world. Also, send the link into the "submit rumors" area of Macrumors so the story makes it to Page 1.

Thanks again Chris.

We're off to the races.
 
Boneheadxan said:
Are the 'Early 2009 MP's' affected aswell? I bought mine in July '09. The post over at Ars mentions the 'Late 2009' models or the Oct '09 models.

I usually have 5 HDD's and a MOTU interface connected at all times. I haven't noticed any drops, but like the post mentions, the fans do not kick in. I usually run Ableton Live/Logic Pro and haven't noticed any performance drops.

If this issue does exist on mine, I should be worried :( I only use iStat for monitoring, and haven't noticed any bumps. I don't use iTunes much, but use Spotify and VLC for movies all the time.


P.S.- What version of iStat allows viewing of Power consumption?
Click to expand...

There's only one real model of 2009 Mac Pro, and it is still on sale now. Apple added the 3.33Ghz option on the Quads, but that's the only change.


iStatMenus allows you to see power consumption.
 
I wrote the author of the article an email and explained that people might get confused by all the EARLY or LATE designations, which I believe are wrong.

So far as I know, ALL Nehalem Mac Pros are called EARLY 2009, and it seems they're all affected, no matter when they were made.
 
Boneheadxan said:
Are the 'Early 2009 MP's' affected aswell? I bought mine in July '09. The post over at Ars mentions the 'Late 2009' models or the Oct '09 models.

I usually have 5 HDD's and a MOTU interface connected at all times. I haven't noticed any drops, but like the post mentions, the fans do not kick in. I usually run Ableton Live/Logic Pro and haven't noticed any performance drops.
Click to expand...


Yes, it affects the early '09 models. I've sent an email to Chris asking him to correct that detail in his article.

If your MOTU interface is firewire or usb, then you most likely are *always* suffering from this bug. If you try benchmarking with the MOTU disconnected, you'll probably see that your machine is faster than you thought it was.


EDIT: Someone else already let Chris know--he's working on fixing it. :)
 
MacNN has a short article too

MacNN has a short article too on this subject, so it looks like the issue is finally getting some attention.
 
I've sent emails to:

Appleinsider.com
MacDailyNews.com
Macworld.com
99.se (Swedens biggest Mac site)

with links to the Ars Technica article.

Let's see what happens..
 
Surely when they were testing this thing before people were able to buy it, Surely they test for things like this? I mean, it's pretty obvious?

I like the idea of doing a test and sending someone here the results, but if you can, please list exactly what needs to be done - as though I was an idiot - start from where to down load the testing software to where to send it.
 
can someone that have experienced the 20% performance drop, try my fix and see if performance goes back or stays the same?


http://groths.se/ABF.zip


(note: should have other OS to access disk and remove kext should something go wrong)
 
ncc1701d said:
Surely when they were testing this thing before people were able to buy it, Surely they test for things like this? I mean, it's pretty obvious?

I like the idea of doing a test and sending someone here the results, but if you can, please list exactly what needs to be done - as though I was an idiot - start from where to down load the testing software to where to send it.
Click to expand...
Download hardware monitor from
Bresink.de
Download Cinebench from
Cinebench for the Mac from apple servers
Run Hardware monitor and put close attention to Heat Sensor Diode of cpua and cpub in case you have an 8 core model and ambient temp
then let your mac pro idle for a few for the temps to stabilize
then fire up cinebench anb run the full set of tests take note of the temps
open iTunes and and play music for about 15 minutes. keep an eye on temps.
Without closing itunes or stoping the music run again the tests of cinebench
Look your performance drop....
 
remove Sleepenabler.kext and boot then, fix will still be applied but without sleep cabability. so you can still check performance.


to remove fix completetly, run Remove from app
 
To me what's exceptionally puzzling is why this problem was not handled as follows:

After many calls reporting this problem into Apple, by users of their most expensive products, Apple did not say "we are aware of the issue, and are looking into a fix; please be patient and check software update" or something similar.

The problem clearly exists, and appears software-based, and hence FIXABLE. Why be so frustratingly bad with customers who are (1) most likely very loyal, long time Apple users, and (2) have purchased your most expensive product? Not everything is about volume. I don't buy that argument. THe Mac Pro is a flagship computing device. It gives Apple cachet. It is not just a concession to a few power users. It gives the company credibility as a sophisticated, hi-tech player. If it just made cheap laptops, it wouldn't have the same image.

Thus a critical software flaw should be addressed ASAP, in my view.

On a side note, I am amazed this is still not on Macrumors.com. :eek:
 
If it is as simple as a software patch then I would like to believe Apple can address this in 10.6.3.

If it was an iMac issue then Apple would have already released the fix in a separate update. But a Mac Pro fix would probably be quietly rolled into an OS Update and never publicly acknowledged by Apple.

Fixing a couple of .kext files can't be that damn hard for Apple. If they can just find time inbetween the iPhone, iPad, and iMac projects. :rolleyes:
 
Cynicalone said:
Hopefully Apple will have to listen now.

I'm still surprised this didn't make the front page of MacRumors.
Click to expand...

You optimist. My hopes of Apple in this are near burned out.
NB. in ars technica article:
"Apple doesn't appear to have acknowledged the problem within its support system. The company did not respond to our request for comment on the issue."

So it may take a lot more than this to shift Apple's attitude. I've been banging at their door for 6 months re. this bug. Sending every sort of evidence, far more than in the ars article. Total stonewall treatment is all I hear from Apple.

What I'm looking forward to is hearing the reason why Apple has been so evasive and un-supportive. But perhaps under the present management we'll have to wait till long after our machines are dead before we hear that.

That said, if Apple pulls its finger out, all this and lots more community effort will be trashed asap.
 
Just for all you nay sayers.

The problem STILL EXISTS.

Tested on a new Mac Pro.

Anyone else that wishes to give me screenshots as above is welcome.

I have two 2.66Ghz Quads here, I need other configs (2.93/3.33 quad, all the octos)
 
so far its on the following sites:

http://gizmodo.com/5463326/nehalem-mac-pros-have-serious-problems-with-audio-processing

http://www.macnn.com/articles/10/02/03/spikes.in.power.heat/

http://arstechnica.com/apple/news/2...ake-20-performance-hit-when-playing-audio.ars

http://9to5mac.com/mac-pro-audio-issues-nehalem-3245934

https://www.macrumors.com/2010/02/0...-and-performance-issues-for-nehalem-mac-pros/

http://www.betalogue.com/2010/01/11/audio-heat/

http://tweetmeme.com/story/523700911/nehalem-mac-pros-take-20-performance-hit-when-playing-audio

http://gadgetblips.dailyradar.com/story/nehalem-mac-pros-have-serious-problems-with-audio/

http://en.wikipedia.org/wiki/Mac_Pro#External_links - have links to the news report there

http://www.xlr8yourmac.com/feedback/2009_mac_pro_cputemps/2009MacPro_cpu_temps.html

Wired - Twittered my Link to Ars
 
Costs

The 2.93GHz CPU of the dual 09 Mac Pro costs £2,125.00 more than the 2.26GHz machine (in the UK, = $3383 in the US). And the latter is 22.8% slower. So as the audio bug slows the machine by 26% on single thread processes (the majority), it looks as if this bug is trashing well over £2k's worth of capital investment (plus the cost of electricity, probably shorter life expectancy of the CPU, and cost of slower work).

Or put another way, your power lust (why you were tempted by the 2.93GHz system) was more than nuked by this audio bug. Expensive wasn't it. Or were you banking on getting first class support for such an expensive machine!
 
I'm glad this is finally getting some traction. :)

I still think the real issue lies in the ICH10R drivers though, as that's where the USB, FW, and PWM controller are physically connected. Power Management is a side-ways approach (CPU), but it's better than nothing at all at this point. ;)
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back