Q: MacBook Pro 2011 17" hard freeze

I ordered the 15" 2.2 i7 and it shipped out today....now I come to find out the machine is probably flawed...disappointing.

Does the 14-day return thing start from the day I receive it or the day I ordered it?

But are you able to reproduce the "Fish GL" problem with the 13" mbp ?

I have none of these issues and have done the tests here (including the aquarium test):

http://mbp-freeze.wikispaces.com/
I ripped one of my DVDs using Handbrake with AppleTV 2 default settings. The movie was nearly 2 hours and 49 minutes--completed the process in 35 minutes.

I maxed out all 8 CPUs using the terminal command. My computer temperature increased to 192 Degrees F. and ran for a while.

Mine was shipped directly from China on 3/1/2011 with the following specs:
17" Antiglare
i7 2.3 GHz Processor
8GB RAM
500GB 7200 RPM Drive

Perhaps because it was a custom order, some of the hardware issues aren't surfacing for me.

Cheers,

Mark

Digimania, did you try to run the Chrome AquariumGL test with the integrated graphic chipset forced ?
1000 fishes rendering.

Adrian J. wrote:
Good to see Engadget picking up the story.

Nice, that should bring a speedy fix

mb852 wrote:
I ordered the 15" 2.2 i7 and it shipped out today....now I come to find out the machine is probably flawed...disappointing.

Does the 14-day return thing start from the day I receive it or the day I ordered it?

I'm pretty sure they base it on delivery date. I had to sign for mine, i think this is why.


I honestly think this issue is software related and will be fixed. The fact that it doesn't seem to happen in windows seems to suggest that it's something in the graphics switching or AMD graphics drivers.

Sorry for some reason even though I set Integrated Only on gfxStatusCard, and it was indeed displaying an "i", it switched to Discrete only (could view it on the menu, didn't appear as a letter, still was showing "i").

Also, I tried again making sure I was on integrated. It's indeed crashing under Google Chrome, but not under WebKit (nightly build). The mouse still moves, so it looks like a purely software crash. With WebKit, I get a solid 30-35FPS with 1000 sharks in Integrated Graphics, without any crash.

Considering the first test that made the GPU switch even though gfxStatusCard was set to Integrated only, this seems clear that gfxStatusCard has a bug. Also it's not nominal working of the MBP with dual GPU, that normally don't support Integrated forcing when using 3D demanding apps.

preppietechie wrote:
In light of what other people are saying about the Chrome aquarium method being unrelated to this issue, that may in fact be true. I attempted to recreate the hang on my MBP 13 (2011) using the "yes > /dev/null &" method and was not able to recreate it. So it appears that the MBP 13's are "safe."

Not necessarily the "yes > /dev/null &" did not cause my computer to freeze but... I have had my MBP freeze when using the Sims 3(when it doesn't freeze it gets super hot).

Also I tried the second method macports, photobooth, and "sudo port install boost" in terminal and it froze my computer within a few seconds.

@ John,

Welp, just returned form another crash thanks to Chrome + swimming fish.

Music continued playing until the end of the song and stopped. I was able to SSH in and attempted to kill all instances of Chrome. I did kill them all successfully, but it didn't get Chrome off my screen or stop the lock-up.

I also sent a reboot command to the laptop. It tried to reboot (even had the spinning wheel), but could not reboot itself fully.

This is obviously not a permanent solution, but for me at least downloading gfxCardStatus (http://codykrieger.com/gfxCardStatus) and switching it to Integrated Only has made the machine a lot more stable.

Of course, now I can't power an external display... but at least it's stopped crashing.

SkyFoxXP,

When I put it into Integration Mode using gfxCardStatus, I get the following message from the Aquarium test:

"It does not appear your computer supports WebGL. Click here for more information."

Cheers,

Mark

That's very consistent. If one of you folks with a 13" computer can confirm, that would be great. But I think that's enough evidence to convince me.

Digimania, this is quite weird because normally you should be able to run the 3D GL test with the integrated graphics... I can't understand this !

Yeah I can replicate exactly this - I noticed this issue not through this thread but by actually doing what was suggested of my own accord - the boost package was a requirement in an installation I was performing with Homebrew (which might be an easier-to-install and less invasive alternative for those hoping to test using this).

I had virtually no apps open other than Terminal and Scrivener, I was on discrete graphics (dynamic switching) and the system locked up. I can replicate this again and again by performing those steps.

This is frustrating. I keep telling myself that a firmware update may fix this, but I worry that there's a larger hardware issue behind it all, hoping against hope that it won't surface and lock me out of my machine, or degrade it's performance a year down the line...

Edit: Incidentally, I must point out that I've been using it for about two weeks with pretty much no issues whatsoever (other than a slight keyboard issue, which Apple dealt with swiftly) and again, as mentioned by others, the system is incredibly cool and very, very quiet.

I run it on integrated graphics most of the time, with discrete graphics heating it somewhat more (though I haven't actually needed discrete graphics particularly, recently).

In the last two weeks it has certainly had its fans sporadically (0 - 1 times a day) spin up heavily, to audible, very high levels, but I had figured that it was just standard heat dissipation until my first crash...

Edit 2: Oh and, incidentally, 2.2GHz model, 8GB RAM. Also, I had the Force Quit issue, and killing apps, including with SIGKILL had no effect, somehow...

Message was edited by: tekacs

Hello. I have the 15" 2011 2.2GHz model and I don't have this problem. I CAN'T reproduce the 'yes > /dev/null' issues with photobooth.