Overheating +FBD Error in Mac Pro. Related?

[coldmember]

Lately, my Mac Pro has started suffering from unrecoverable FBD errors and required restarting daily. I had to restart it both yesterday evening when I came home from work and then again this morning when I toke up.

I was only running BOINC on both occassions.

In January, I upgraded the RAM from the stock 1GB with 2 1GB hygynx (sp?) FBDIMMS. It passed 5 REMBER loops after the install and I've not had a problem with it until this past week.

Last night, I ran REMBER again with no errors detected.

When I woke this morning I found the restart message again. This time, I turned the Mac off and removed the hygynx chips. They were incredibly hot to the touch.

So I'm wondering if that may have something to do with the errors I'm getting.

Currently, it's back in the stock 1 GB confirguration until I figure out the next steps.

The recent cold spell here has prompted me to turn the heat up in the house....ergo, higher ambient temperatures. Will this have a serious affect on the Mac's internal temps? Is there anything else I should consider?

Thanks!

 

[netnothing]

That's interesting because I was just about to post another topic related to this. I got my Mac Pro back mid-march, came with the stock 1GB RAM (2x512). I can only afford right now to order the 2GB options to bring my total RAM to 3GB.

I first ordered from OWC, 2x1GB DIMMS. After installing, I was noticing under heavy applications, System Profiler was giving me ECC errors.

So I had OWC ship me new sticks. Well the new sticks didn't do much better, so I sent them back. During this whole process with OWC I had 1 kernel panic and lots of ECC Errors. I tended to think that the ECC errors were due to the heat. The OWC DIMMS would get HOT under load, over 160F.

Now, I ordered the 2x1GB DIMMS from Crucial. Been running those for a few weeks. No ECC errors ever seen in the System Profiler, but I have gotten 2 kernel panics the past 2 days. So, I'm returning those and Crucial are sending me out new DIMMS.

I'm beginning to wonder what is going on. The kernel panic I had yesterday, I was just doing some Dreamweaver work...not anything intensive. This morning I decided to run Rember, and around test 5, I got a kernel panic.

All of my kernel panic's have been the FBD errors.

I've run the Apple Hardware test multiple times and never had it report an error. Also, during the OWC testing, I did move the Apple RAM to the lower riser to make sure the riser wasn't bad, and I still got the ECC Errors.

I can at times see where the heat might be causing problems...but yesterday it wasn't hot, and I wasn't doing anything intensive.

I guess the questions I have for the experts out there are:

1) Are the FBD errors kernel panics, strictly related to RAM?
2) If so, do people think that it could be related to heat?
3) Are people out there running smcFanControl to ALWAYS keep their fans at a higher level?

I've spent a lot of money on this Mac Pro...and I am starting to get worried that it's not going to be reliable for my day to day work. I've moved everything over and I NEED this machine to be rock solid.

Any thoughts?

coldmember:
Can you post your temps? Both mem temps and Ambient temps? Are you running Temperature Monitor? http://www.bresink.com/osx/TemperatureMonitor.html


-Kevin

 

[coldmember]

Thanks for the link.

Ambient temps are now a much cooler @ 27C now that the SO has turned down the heat. Earlier, they were, at a guess, 40C. This is due not only to the radiator, but also 3 other computers (including a Dell PowerEdge 2300 server) running. The computers, alongside the two Dell 24" widescreen displays, contribute quite a bit of heat to a small room.

Memory temperature is 71-74F at modules A1 and A2. CPU core temps range from 55 -59 C while Memory Bank temps are ~ 46 C in Bank A (active) and 40 C (no RAM).

The Mac is only running BOINC at the moment. If I don't get any kernel panics overnight, I'll reinstall the other RAM and see if the cooler ambient temps help.

I'll post temp updates tomorrow after it's been running all day.

 

[Kosh66]

Yes, heat could definitely be the issue. FBDIMMs run hotter (I believe it's due to a controller chip on them) and if they aren't properly cooled will cause ECC errors. Hopefully the cooler room temps will solve your problem...

 

[CanadaRAM]

As said, heat = more memory errors. If your machine is sitting in an environment above room temperature (even 27C (80F) is on the warm side -- 40C (104F) is way overboard) then the machine is not going to be able to move the heat out fast enough.

Apple's official environmental operating specs are 50° to 95° F (10° to 35° C)

 

[netnothing]

So why hasn't Apple addressed the Fan issue on the Mac Pros? I mean, c'mon....why do I have to manually kick up the fan levels. I can run Rember through 10 loops....get the DIMMS up to 170F+, and the fans never kick in.

I'm getting my second set of memory from Crucial tomorrow. I'm hoping that the 2 kernel panics I had yesterday were from the other RAM.

At this point I'm nervous using my system because the way OS X works. It seems far to easy to kernel panic.

Does anyone know....is the FBD kernel panic strictly related to bad memory, memory that gets too hot, or can an application cause this error as well?

-Kevin

 

[miniConvert]

I bought extra Apple RAM from an Apple Store, because I have more money than sense, with the heatsinks included and then fitted it myself. I took my machine from 1GB to 4GB RAM.

No ECC errors, no kernel panics, even when the temperatures in my office soar or I'm cooking the ATI X1900XT with WoW

You could use smcFanControl to turn up the fans from their standard 500rpm? You can add a few hundred without a noticable noise increase.

 

[coldmember]

As said, heat = more memory errors. If your machine is sitting in an environment above room temperature (even 27C (80F) is on the warm side -- 40C (104F) is way overboard) then the machine is not going to be able to move the heat out fast enough.

Apple's official environmental operating specs are 50° to 95° F (10° to 35° C)

I'm horrible at metric conversion. Ambient temp was probably pushing 90F, though.

I re-installed the aftermarket RAM this AM before work. It has the same heat sinks as the stock RAM, btw. We'll see how the lower ambient temps help.

 

[coldmember]

After running all day with an ambient temp of 73-75 F, Memory Modules A1 and A2 (2 x 512 MB Apple RAM) each ranged between 150-154F.

Module B1 and B2 (2 x 1GB aftermarket with Apple type heatsinks) ranged from 156-159F and 158F -161F, respectively.

Is this on par with what others are seeing?

 

[netnothing]

After running all day with an ambient temp of 73-75 F, Memory Modules A1 and A2 (2 x 512 MB Apple RAM) each ranged between 150-154F.

Module B1 and B2 (2 x 1GB aftermarket with Apple type heatsinks) ranged from 156-159F and 158F -161F, respectively.

Is this on par with what others are seeing?

I averaged about 70F all day today, and my stock Apple DIMMS topped out at 130F/122F.

My NEW Cruical DIMMS are working great, they topped out at 127F/131F.

What is your setup like? Do you have ample space behind the Mac for airflow?

-Kevin

 

[coldmember]

I averaged about 70F all day today, and my stock Apple DIMMS topped out at 130F/122F.

My NEW Cruical DIMMS are working great, they topped out at 127F/131F.

What is your setup like? Do you have ample space behind the Mac for airflow?

-Kevin

It's beneath my desk and away from the wall. There's no enclosure there, the entire area is open.

For clarification, are those the temps of your memory modules or memory risers? I've noticed that my risers are 10-15 degrees cooler than the modules themselves.

 

[netnothing]

It's beneath my desk and away from the wall. There's no enclosure there, the entire area is open.

For clarification, are those the temps of your memory modules or memory risers? I've noticed that my risers are 10-15 degrees cooler than the modules themselves.

Those are for the modules themselves. The risers are running 90-99F right now with an ambient temp of 71.6F.

I have noticed that these 2 new sticks of Crucial RAM are running WAY cooler than the last two, or an of the OWC DIMMS I got. Can't explain it, but I'll take it.

-Kevin