Hi!
I'm new to this forum and I want to tell you my cMP story and ask you for advice on some issues I still have.
I bought a cMP 4,1 at the tail end of last year that had its firmware upgraded by the previous owner already. This is my first Mac ever and I'm not sure I would want another model, maybe a trash can if the prices ever drop below 120 Euros.
Since everything worked great I bought two delidded X5680s, 128 GB of RAM and an NVMe with a PCIe adapter. I installed Linux on it and all the other parts without issue.
Everything went smoothly until a few days after I installed the new CPUs. The device shutdown while it was idling away for an hour and flashed both of the OVTMP_CPU lights constanly every few seconds and would do nothing when pressing the power button.
After some research it turned out that apparently early, early 2009 daughterboards could die from 130W CPUs. That is a fact that is barely mentioned, if at all in most guides. There was a footnote in the Definitive Classic Mac Pro Upgrade guide, but I must have already forgotten about that when I bought the CPUs.
So I checked the tray's serial number and it starts with J5908..., 8th week of February. Obvious conclusion: I killed the tray.
However that made no sense to me, after all the new CPUs survived a half hour stress test before, so why are they failing now at idle?
So I put the original e5520s in again and prayed: to no avail, the red blinking ligths of death were still after my soul and I can still see their ghostly afterimage every time I plug in a Mac Pro. And I'm always so relieved when they don't light up more than the one time they're supposed to.
Later that same night I had the genius idea of removing CPU B and that actually worked flawlessly. Mbpfan (a Linux fan control software for Macs)
had no trouble keeping the fan down that would normally go full blast when running a single CPU in a dual tray.
Happy that I still had half a working CPU tray I of course tried to get it working again by reinstalling the second e5520. That also worked, but one RAM slot was inactive. I took that as a pyrrhic victory and went ahead and reinstalled the PCIe devices I wanted to use. And those horrible lights that shall not be named returned without me touching the CPUs!
That was It: After reinstalling both CPUs twice more I was sure the tray had become completely unstable from the 130w Xeons, so I took the opportunity to buy another 4,1, and sell the old one for parts. Dual CPU tray prices are ridiculous compared to full systems so it doesn't make sense to buy those as spare parts.
New Mac who dis?:
The new 4,1 was next to unmolested since it was employed in an agency. Immidiately I had to update the firmware because apparently the original firmware didn't support booting from my NVMe with Linux on it. It took me a while, but with the Snow Leopard dmg and Transmac on a Windows system I finally got OS X on that Mac. Even as a passionate Linux ricer I have to say: Snow Leopard looked great. But I had to leave it for El Capitan to crossflash and for High Sierra to get the Firmware to 144.0.0.0. (didn't install Mojave, just patched the firmware).
So I put my NVMe in my new "5,1" and guess who was back? Go figure, it was red and it blinked and somehow spread like a virus to a different machine to take my first born son.
The culprit was easily found, both the CPU trays worked in both MPs with the High Sierra drive. IT WAS THE PCIE ADAPTER THAT WAS BAD. The SSD work
ed with a different adapter, however. I guess never buy the last one of the model they have in stock for a discount. RIP to those 8 Euros. But when a machine tells you there is an issue with it's processors and you swapped them less than a week prior, and you just found out about a dangerous incompatability would you have suspected the SSD drive? I don't think so.
In conclusion: I now have two perfectly working 4,1 > 5,1 cMPs as far as I'm aware, but the story hasn't ended, yet. I still have to swap the e5520s for something more potent in the one I want to use. I still have the X5680s, but they are not going in the J5908... tray again. I played with fire there and I won't again. The tray of the newer machine has the serial number: *C07012405NBDCVHA9*, so it was refurbed by Apple and it's a 639-1060 tray. Can anyone say for (close to) certain wether it's safe to use 130W processors in that board? Or should I just get the x5675s and sell the others?
And then there is the bootROM problematic, I've combed through some threads but I still don't understand it fully. Do I even have to worry about t
his if I only use Linux? Because it sounded like it's an issue with multibooting different OSes, especially old versions of Mac OS and Windows in
UEFI mode.
I still have the drive with High Sierra for diagnostic purposes but I have little interest in using Mac Os. I would however like to run Windows on it, ideally without bootcamp for a stupid joke:
I'm planning on doing a Pixla mod and buying another of the R9 280x that came with the first MP to run in Crossfire X and see wether gaming with 2 CPUs and 2 GPUS is a bad idea or a very bad idea. That would be a fun idea to bring to a LAN. I'm probably going to sell one of the MPs after I'm done with the mods on the other.
So it would be really nice if somebody could tell me what I can do to stay on the the safe side with my Mac Pros, especially with the bootROM. I don't want to see those eldritch crimson illuminations ever again, I don't want to fry 180 Euro daughterboards with 30 Euro CPUs and I don't want my device to turn into a doorstop because some flash chip on the motherboard with a ticking firmware bomb on it decides that this is "the day".
Thank you all for reading my long post and offering your help!
It really has been quite the experience with these computers.
I'm new to this forum and I want to tell you my cMP story and ask you for advice on some issues I still have.
I bought a cMP 4,1 at the tail end of last year that had its firmware upgraded by the previous owner already. This is my first Mac ever and I'm not sure I would want another model, maybe a trash can if the prices ever drop below 120 Euros.
Since everything worked great I bought two delidded X5680s, 128 GB of RAM and an NVMe with a PCIe adapter. I installed Linux on it and all the other parts without issue.
Everything went smoothly until a few days after I installed the new CPUs. The device shutdown while it was idling away for an hour and flashed both of the OVTMP_CPU lights constanly every few seconds and would do nothing when pressing the power button.
After some research it turned out that apparently early, early 2009 daughterboards could die from 130W CPUs. That is a fact that is barely mentioned, if at all in most guides. There was a footnote in the Definitive Classic Mac Pro Upgrade guide, but I must have already forgotten about that when I bought the CPUs.
So I checked the tray's serial number and it starts with J5908..., 8th week of February. Obvious conclusion: I killed the tray.
However that made no sense to me, after all the new CPUs survived a half hour stress test before, so why are they failing now at idle?
So I put the original e5520s in again and prayed: to no avail, the red blinking ligths of death were still after my soul and I can still see their ghostly afterimage every time I plug in a Mac Pro. And I'm always so relieved when they don't light up more than the one time they're supposed to.
Later that same night I had the genius idea of removing CPU B and that actually worked flawlessly. Mbpfan (a Linux fan control software for Macs)
had no trouble keeping the fan down that would normally go full blast when running a single CPU in a dual tray.
Happy that I still had half a working CPU tray I of course tried to get it working again by reinstalling the second e5520. That also worked, but one RAM slot was inactive. I took that as a pyrrhic victory and went ahead and reinstalled the PCIe devices I wanted to use. And those horrible lights that shall not be named returned without me touching the CPUs!
That was It: After reinstalling both CPUs twice more I was sure the tray had become completely unstable from the 130w Xeons, so I took the opportunity to buy another 4,1, and sell the old one for parts. Dual CPU tray prices are ridiculous compared to full systems so it doesn't make sense to buy those as spare parts.
New Mac who dis?:
The new 4,1 was next to unmolested since it was employed in an agency. Immidiately I had to update the firmware because apparently the original firmware didn't support booting from my NVMe with Linux on it. It took me a while, but with the Snow Leopard dmg and Transmac on a Windows system I finally got OS X on that Mac. Even as a passionate Linux ricer I have to say: Snow Leopard looked great. But I had to leave it for El Capitan to crossflash and for High Sierra to get the Firmware to 144.0.0.0. (didn't install Mojave, just patched the firmware).
So I put my NVMe in my new "5,1" and guess who was back? Go figure, it was red and it blinked and somehow spread like a virus to a different machine to take my first born son.
The culprit was easily found, both the CPU trays worked in both MPs with the High Sierra drive. IT WAS THE PCIE ADAPTER THAT WAS BAD. The SSD work
ed with a different adapter, however. I guess never buy the last one of the model they have in stock for a discount. RIP to those 8 Euros. But when a machine tells you there is an issue with it's processors and you swapped them less than a week prior, and you just found out about a dangerous incompatability would you have suspected the SSD drive? I don't think so.
In conclusion: I now have two perfectly working 4,1 > 5,1 cMPs as far as I'm aware, but the story hasn't ended, yet. I still have to swap the e5520s for something more potent in the one I want to use. I still have the X5680s, but they are not going in the J5908... tray again. I played with fire there and I won't again. The tray of the newer machine has the serial number: *C07012405NBDCVHA9*, so it was refurbed by Apple and it's a 639-1060 tray. Can anyone say for (close to) certain wether it's safe to use 130W processors in that board? Or should I just get the x5675s and sell the others?
And then there is the bootROM problematic, I've combed through some threads but I still don't understand it fully. Do I even have to worry about t
his if I only use Linux? Because it sounded like it's an issue with multibooting different OSes, especially old versions of Mac OS and Windows in
UEFI mode.
I still have the drive with High Sierra for diagnostic purposes but I have little interest in using Mac Os. I would however like to run Windows on it, ideally without bootcamp for a stupid joke:
I'm planning on doing a Pixla mod and buying another of the R9 280x that came with the first MP to run in Crossfire X and see wether gaming with 2 CPUs and 2 GPUS is a bad idea or a very bad idea. That would be a fun idea to bring to a LAN. I'm probably going to sell one of the MPs after I'm done with the mods on the other.
So it would be really nice if somebody could tell me what I can do to stay on the the safe side with my Mac Pros, especially with the bootROM. I don't want to see those eldritch crimson illuminations ever again, I don't want to fry 180 Euro daughterboards with 30 Euro CPUs and I don't want my device to turn into a doorstop because some flash chip on the motherboard with a ticking firmware bomb on it decides that this is "the day".
Thank you all for reading my long post and offering your help!
It really has been quite the experience with these computers.
