Currently Being ModeratedApr 28, 2012 5:06 PM (in response to Bill Strohm)
I'd try turning off CPU1
You might try this in Terminal or Single User Mode if you don't have CHUD tools...
sudo nvram boot-args="cpus=1"
The cpus boot argument can range from 1 up to the total number of processor cores on your system.
And thanks to old comm guy...
This can also be done in open firmware as well (boot with cmd-option-O-F held down.
setenv boot-args cpus=1
Currently Being ModeratedApr 29, 2012 2:45 AM (in response to Bill Strohm)
I've just spent fighting the exact same boot symptoms + almost immediate (few minutes usually) crashing when booted into Leopard for some days. It could be the CPUs, but I'd place my bets on the well-known logic board BGA solder issue. I had an original 2003 G5 logic board and a June 2004 G5 tower to play with. I've run tests on the CPUs for hours and they don't crash, but the logic board test as well as anything in the OS that transfers lots of data through the U3 chip quickly end up in a lockup. For some people, the hair dryer trick (google it) helps, but it doesn't fix anything really.
The discouraging fact is that I think most of these logic boards are starting to be in the same condition; they just can't take years of power on-off heating/cooling cycles. The best replacement part would probably come from a machine that has had very little use or been running 24/7.
btw. AFAIK the 2.7 GHz model can't boot up with one CPU only.
Currently Being ModeratedMay 1, 2012 9:40 AM (in response to Heikki Lindholm)
First, I apologize for the loss in formatting of my original message. I was using an old desktop G3 running OS 9.2.2 and "Classilla," and for some reason all my original formatting was lost when sent. Now I am replying on my wife's Windows computer, since no text cursor appears in the "Reply" window of this forum when using the G3.
Next, it is true that if one CPU is shut off on my G5, during restart both CPUs are turned on again. Here is an update on my problem:
When I retrieved the G5 from the Apple Store after their investigation, I found they had (inadverntently?) left a CD in its drive. That turned out to be a copy of Apple System Diagnostics, which I have been trying to use to find an error. The disc contains two volumes, plus a "Read Me" file and a "User Manual" in PDF format. I read the instructions and attempted to follow them.
Per the instructions, I removed the Ethernet and one of the two
Currently Being ModeratedMay 1, 2012 10:15 AM (in response to Bill Strohm)
I tried to edit the above post, but the edit was not accepted by the forum . Here is some addiional information. Post continues as follows:
... USB cables. I reset the PMU, reset the PRAM, and booted with "C" depressed. The Apple System Diagnostics opened and displayed the "OF" (Open Firmware I guess) window. I ran all 129 tests, which passed with no errors. (According to the User Manual, that test does not require Mac OS X at all.) Then per the instructions, I restarted using the Option key, and chose the "OS" System Diagnostics volume.
The results to date have been unsuccessful. They vary from an Apple logo with no spinning "gear" to a spinning gear that freezes, to a "Mac OS" progress bar that freezes, to an actual "OS" diagnostics window appearing but frozen, to a test actually beginning but freezing before either a "Pass" or a "Fail" showing up. At this point I am stuck.
If this problem is indeed a BGA solder issue, I have no idea how to further isolate it. Is it the PMU chip? The PRAM chip? A CPU?
Any more ideas? Thanks for your previous help!
Currently Being ModeratedMay 1, 2012 11:26 AM (in response to Bill Strohm)
Most probably the solder issue happens in the U3 system controller which is situated on the backside of the logic board, with a large heatsink assembly. They should have cooled it better. Even on 2 GHz machines it heated up to 80 C quite easily. There's an IBM document  boasting how well they designed the thermal pumping endurance of the CPUs (also BGAs) - I don't know whether they applied that bit of engineering to the U3 controller as well..
With the older G5 generation as well as the latest one, it's possible to run the machine with only one CPU attached, and see whether it's one the CPUs that's starting to give up. I don't think the 2.7 model will start up like that, unfortunately.
You could fire up the OF test from the ASD CD and let it run the logic board test (select individual tests) for like 50 runs (loop count) or so and see if it locks up. The problem with the CD is that the fans go full rpm, making the cooling much better than during boot up or in the OS. One of my broken logic boards hung up during the ASD test deterministically and the other only in the OS or during boot up.
You could also loop the CPU tests from the ASD, but they don't seem to be a stress test, but rather just test computational correctness. Looping them probably won't put much heat load on the CPUs. If you can boot into OS, using some small stress test that fits in the CPU cache (small ffts etc..) could be used to verify the CPUs. The ASD memory test should be ok to test the memory buses, but those also go through the U3 chip.
Currently Being ModeratedMay 12, 2012 10:23 AM (in response to Bill Strohm)
First, thanks for all the advice above. Second, I apologize for this tardy response; my G5 has been in the Apple Store repair shop for the last week, and my old desktop G3 running Mac OS 9.2.2 has not allowed me to reply to this thread (for some reason clicking "Reply" does nothing).
This is a final update on my repaired G5:
Problem was isolated to the dual-CPU assembly. This unit includes the LCS, which had no problem, but the whole shebang is replaced as a single piece, price $909. Add $70 tax + $39 labor and the total cost was $1018. Yep, a lot, but I had paid $4K for the computer back in 2005, so now It's a $5K computer and I can still use all my software. Disadvantages of just buying a Mac Pro: No Classic to play all my old video games, MS Office 2004 wouldn't run and would have to be replaced, plus all my other unique (non-Universal apps) would not run.
According to Marcel Bresink's Hardware Monitor app, the new CPU chips both use less power than the old ones, and, probably as a result, run a little cooler.
Looks like there was no problem with the Logic Board, which the Apple Diagnostics (OF) suggested when run on the old CPU assembly. I did run 50 loops on the non-RAM OF tests, with no errors. I guess all the Logic Board ICs are OK; everything now seems to work just fine.
Again, thanks to the forum for your suggestions!