You can make a difference in the Apple Support Community!

When you sign up with your Apple Account, you can provide valuable feedback to other community members by upvoting helpful replies and User Tips.

Looks like no one’s replied in a while. To start the conversation again, simply ask a new question.

Several sudden shutdowns on new iMac

My new iMac had several sudden shutdowns in the last 30 minutes.
Just normal use, open were mail, safari temp. monitor.
Here is a picture of the log just before crashing.
http://i767.photobucket.com/albums/xx320/mariner4/Schermafbeelding2010-08-29om17 1649.png

Hope someone can make out something.

peterwillem

macbook pro 4,1; iMac 27" i7, Mac OS X (10.6.4)

Posted on Aug 29, 2010 9:36 AM

Reply
136 replies

Nov 26, 2010 10:46 PM in response to peterwillem

My computer isn't shutting down (well once) but I have the normal factory installed 2x2gigs. I went out and got 2x4gigs (Kingston) for a total of twelve. I wasn't noticing any speed improvements in photoshop CS5 so for a goof I clicked on "About this mac" and it told me I only had 6 gigs of ram. I clicked on more info and then memory and it told me that two of my banks were empty! It showed one of the 2gigs and one of the four gigs. Did a bit of researched and put 4's on the stand side and the 2's on the screen size. Presto, I clicked on "about this mac" and it said 12 gigs of ram. Next time I went to my computer after it had been asleep I checked it again for a goof and it said 6 gigs of ram!! It shows again that two of my banks are empty the other two have a 2 gig and a four gig. What the Fu@k is going on? Any ideas?

Dec 1, 2010 12:07 PM in response to peterwillem

I've went down to 14GB 12 of OWC's NuRam & 1-2GB stick of the original ram supplied by Apple on November 15th, so far not as bad but not totally stable. No random shut downs however I did have one lockup which forced me to hold the power button down to shut it down. The system rebooted normally but the display was asleep when it was frozen and even the old 3 finger salute wouldn't phase it.
The only other strange behavior later on November 20th was an odd message telling me that my "Your Mac start up disc has no more space available for application memory." A 3rd party program, SnowTape was running but wasn't responding and the other open applications were Mail, Safari and Finder. I only opened Little Snapper to do a screen capture that included the message along with the info from iStat.
Here's what I captured from the iStat drop down:
Wired 674MB
Active 8.87GB
Inactive 4.38GB
Free 79MB
Page ins: 78,895
Page outs: 5,169,368
Swap: 59.00 GB
Snow tape 12.07 GB(!!)
Windows Server 110MB
Mail 80MB
Helper 55MB
mds 48MB
BTW I'm no tech and I'm not woking for Apple for free either but aren't those "Page Outs" freaking HUGE?! Seems like I recall reading that you should have more Page Ins than Page Outs normally or something is way wrong. Memory Leak? No or incorrect Garbage Collection in OS X or in other apps or perhaps something completely different? John Cleese?
I'm curious if it's possible that my migration from a 32 bit Intel iMac w/a Core Duo could have anything to do with it since I used XSlimmer quite a bit to trim the PPC code and some Intel 64 bit long before I purchased my new Core i7 iMac? Another reason I brought that up is an unrelated topic on BluRay ripping where they recommend I reinstall OS X.

Dec 2, 2010 2:51 AM in response to photoman67

I went out and got 2x4gigs (Kingston) for a total of twelve.


What is really wrong nobody knows. Even Apple doesn't.
I only can tell you that this iMac is very particular with ram. Only Apple original or Samsung original factory seems good enough. Kingston, although good ram, not good enough for this iMac.
OWC's own ram is ok as long as you do not use more than 12 G.



peterwillem

Dec 2, 2010 5:00 PM in response to peterwillem

Bonjour,

Je pense que ce problème provient du nouveau hardware du iMac. En effet sur les nouveaux quad core i5 et i7 le Front side bus FSB a disparu au profit de contrôleur intégrés aux processeurs…. Ceux ci assurent maintenant les échanges directs entre les CPU et la Ram (Quick path architecture voir lien suivant:
http://www.appleinsider.com/articles/09/02/02/nextgen_mac_pro_processors_could_arrive_march29.html )

Ceci est un changement hardware majeur !!! Je pense donc que avec cette architecture les processeurs soient maintenant très sensibles à la charge représenté par les barrettes mémoire RAM. Cette charge peut varier selon la fabrication du board des barrettes, le nombre et le type de puce sur les barrettes ect… Donc plus il y a de barrettes plus il y a de puces mémoire et plus la charge est élevé. Je pense que la capacité des contrôleurs intégrés aux puces soit trop limite pour contrôler 16 Gig de mémoire plus résistive que celles de samsung et Hynix . Ceci expliquerait aussi que dégagement de chaleur est plus élevé avec une quantité de mémoire élevé. (processeurs surchargés…) Aussi il faut savoir que à ces fréquences élevé l'appariement des impédances des mémoires au circuit devient aussi très critique et que celle ci varie pour pour chaque barrette de fabrication identique… Donc un autre ensemble de barrettes identique peut réussir ou un autre ensemble identique a échoué…

Finalement j'ai bien peur que ce problème ne puisse être résolu par une révision de Firmware ou autre.

Salutations !

mechanical translation follow;

Hello,

I think that this problem results from the new hardware of the iMac. Indeed on new quad bike core i5 and i7 the Front side bus FSB disappeared for the benefit of controller integrated into processors. Those this assure now the direct exchanges between the CPU and the Ram (Quick path structures , see following link:
Http: // www.appleinsider.com/articles/09/02/02/next gen_mac_pro_processors_couldarriv e march29.html)

This is a change major hardware!!! I thus think that with this architecture processors are now very perceptible to the load represented by memory pin RAM. This load can vary according to the manufacturing of the board of pins, the number and the type of (chip) on pins ect … Thus more there is of pins more there is of chips memory and more the load is raised. I think that the capacity of the controllers integrated into CPUs is too limit(borderline) to control 16 Gig of memory more résistive than those of samsung and Hynix. This would also explain that release of heat is more raised with a quantity of high memory. ( Overloaded processors) Also it is necessary to know that with these frequencies raised the matching of the impedances of memory to the circuit becomes so very critical and that that this varies for for every pin of identical manufacturing … Thus another identical set pins can succeed or the another identical set failed …

Finally I am well afraid that this problem cannot be resolved by a revision of Firmware or other.

Greetings!

Dec 3, 2010 4:14 AM in response to champagne999

Merci Champagne,

Your post makes sense. I am not a technician so I don't follow you all the way.
If I understand you correctly it is because of the lack of a FSB and therefore the ram being directly via pins connected to the cpu there is often a kind of overload when using full 16 G ram capacity . All this because of new hardware (iCore cpu) Apple put in this machine.
And the number of pins and chips play a role. Maybe Apple should have tested it better?
But then I have this question: Some type of ram (Apple, Samsung, Hynix) do not have these problems but other kind of ram ( OWC, Crucial, Nu Corsair etc..) do have these problems. But all the rams mentioned here have the same amount of pins and the same amount of chips, so how come that some have problems and some don't?

Anyhow it looks like this problem can not be solved, unless you know which ram will work!
Apple should test the various kind of ram and make a list which ram will work with 16 G.
That is if there is no other solution.



peterwillem

Dec 3, 2010 9:39 AM in response to champagne999

Hi,
excuse the bad english translation... please replace the word _pins by module_ ! Then, one memory module by memory slot is possible, there are many memory chip on one memory module, there are different material for memory module board, there are different memory chip for the same memory chip capacity... etc

The impedance is the electric resistance caused by high frequencies.

I am sure that Apple know this problems...But she has no advantage to announce it... In most she could announce that only his (very expensive) memory module is of sufficient quality for a maximal capacity of 16 Gig!

Greetings!

Dec 7, 2010 12:08 AM in response to peterwillem

I was contacted by phone early Monday morning by an Apple representative and was very pleased with our conversation in assisting me and everyone in this matter. I'm happily complying with his request in providing as much information as I can to aid him, the Apple Engineers and technicians in solving this matter.

PS
If I was offered a new iMac w/16GB of Apple installed Ram, as another forum member claimed, in trade for mine I'd do it in a heartbeat!

Of course IF that would fix the issue. However, even though it may not matter all that much if it works, would Apple be installing the same ram as the factory? I cannot see where the Apple ram is made by Samsung or at least it's not labeled as Samsung.

Dec 7, 2010 5:48 PM in response to champagne999

Thank you Champagne999 for your message. (je parle français mais je vais utiliser l'anglais dans ce forum)

I have the same problem.

ISSUE: ________________________________________________
iMac shuts down WITHOUT warning

INFORMATION: _______________________________________
• 27" 2.93 GHz/ Intel Quad Core i7 / 12GB / 2TB / OS 10.6.5
(12 GB, 4GB from apple / 8GB from G.Skill memory)
• All firmware updates have been applied according to software update
• 3 external drives: 2-TB Netgear stora / 2-TB USB 2.0 Fantom / 2-TB USB 2.0 WD

I have the same problem. It's a nightmare. I bought my iMac (27'' i7 2.9Ghz 12GB 2TB) on august 2010 and it has shut off randomly without warning 17 times in 14 weeks. Sometime, i was surfing on the internet, sometime I was using iMovie or iPhoto. I went 4 times at apple store but they have never repaired it because they can't reproduce the problem. Since I have my iMac, it was on reparation over 17 days on 120.

They don't want to replace my narcoleptic iMac with a new one. It's my first imac and begin to regret to have switched from windows. I already had a problem with a Dell computer. I called and the day after I received a new computer. I don't understand why Apple don't want to replace my 3 months iMac. I loose my time to explain, make backup, pack, go to store, unpack, format, install.They asked me to plug my computer in a new room! I have done but it shuts off. Buy a battery backup! Disconnect all USB devices. What's next? In fact, they don't know what to do.

In fact, Apple don't know what to do. On dec 2, they changed the power supply. 5 days after, my imac shut down.

If I understand you, there is nothing to do with this problem. On this forum, we are 50 to have this problem. May be it's not a pandemic issue.

Dec 7, 2010 7:15 PM in response to champagne999

Although not a Mac site nor post it's an interesting read on the Core i7 and it's integrated memory controller problems. http://www.tomshardware.com/forum/274636-28-core-broken-memory-controller

A couple of quick snipps/tidbits from this site:

I had 12gb installed on my i7 build, one day it decided to say 8gb
I did all the troubleshooting possible and the most I get was 10gb when I had 12gb on!! Both the BIOS and the OS said 10gb. First I thought it was the mobo, then I test every memory individually in every memory slot and they all work. I contact ASUS and they told me to do CPUZ (dumb me, I thought that the BIOS said everything...) and I saw the 12gb, 2gb in every slot. In the end, I contacted Intel to get a replacement and just sent in my 920 today. Have this happened to anyone???

I actually have heard of this before, several times. I do not have a 920 so I don't know a ton about it, but in some cases raising the voltage to the memory controller helped (in others it didn't though, so your problem may very well have been a defective component).

You my friend are lucky! because i just got from the mail my new core i7 replaced by the warranty. LOL i had C0 stepping and now they sent me the newer D0 stepping models.

Well I just got back from the processor store and my i7 is working fine >.<S-SPEC: SLBEJ yay!!!!!!!!

Another one & another tid-bit: <a class="jive-link-external-small" href="http://">http://www.tomshardware.com/forum/259931-28-could-defective-memo ry-controller


One observation I want to point out after reading multiple threads on this issue at multiple sites:

A commonality is that the overwhelming majority of folks experiencing these memory set-up issues are running both i7 920's w/C0 stepping and some variant of the ASUS P6T family of non-workstation boards, be it the P6T, P6TD, or the latest P6TDv2.

I am also experiencing these issues using a P6TDv2/i7 920 C0/Dominator 1600 matched set DIMMs.

After two weeks of fiddling around with different memory and every setting known to man, I am also going to RMA the CPU and see what happens. Hopefully, I will get a D0 chip.

I can't help but wondering, however, why this is happening to us poor slobs that chose these boards. All things considered, there appears to be something different in the memory management features when compared to the same class of MSI and EVGA boards.

It's all speculation at this point and Apple is working on this but they cannot say too much w/o offering a solution or fix, yah know.

Dec 25, 2010 12:23 PM in response to peterwillem

OK - add me to the list of surprise shutdowns. Brief history:

10/24/10 iMac i7 purchased from Apple online with 4 GB memory installed.
11/7/10 installed 16GB memory from OWC.
Late November, 2010 iMac just shut down by itself - just reading something on Safari. I did not note the exact date because I thought this was just an isolated occurrence.
11/30/10 Shutdown at 11:20 AM - just reading something on Safari.
12/1/10 Shutdown at 6:10 PM - again, just Safari.
12/20/10 Shutdown at 8:22 AM - again, just Safari.
12/24/10 OWC sent me new 16GB of memory and I installed it at 4:00 PM.
12/25/10 Shutdown at 12:03 PM - again, just Safari.

These shutdowns are quiet with sometimes just a very soft pop - no messages of any kind. I have display sleeping at 15 minutes but set at never turning off computer.

On the original OWC 16 GB, I have run Apple's extended hardware test and a total of 32 passes on the Rember memory test - no problems. I also ran 5 passes on the new 16 GB with no problems on 12/24/10.

As you can note from my timeline, this is truly random. Also note that I am in my 60's and very protective of my computer.

One great thing about Apple Discussion is that I am not alone. Not great is that an expensive computer fails.

If somehow I missed a solution other than just use 12 GB, please let me know.

Thanks.

Dec 26, 2010 12:04 PM in response to peterwillem

Update 2:

As I posted previously, I performed the Rember memory test which was suggested by OWC.

All three times (total 32 passes) I did this test on the first OWC 16 GB memory and the 5 passes on the replacement OWC memory, I received the following message: "Attempting memory lock... WARNING: Testing with unlocked memory may be slower and less reliable ERROR: Memory lock failed - reason unknown."

I performed the Rember test on the original 4 GB Samsung memory after I re-installed it on my iMac and received the following message: "Attempting memory lock... locked successfully"

On the OWC memory, each pass took 1 hour. On the Samsung original memory, a pass took 5 minutes.

I am not really sure what memory lock is, but it appears that the i7 is not compatible with OWC memory.

I will relate my findings to OWC tomorrow 12/27

Several sudden shutdowns on new iMac

Welcome to Apple Support Community
A forum where Apple customers help each other with their products. Get started with your Apple Account.