Precision T5500 Green Power, solid 3, 4

Question

I acquired a dual CPU T5500 months ago and set about upgrading it as much as I could. First thing was a 1TB SSD (up from 250GB SSD & 500GB HDD) with Win 10 Professional and 24GB SAMSUNG RAM modules (6 x 4GB ECC) (up from 12GB (6 x 2GB ECC)). So far so good. Next upgrade was installing an NVIDIA GTX 1080Ti video card to replace the NVIDIA NVS 420. Again so far, so good. The last upgrade was installing dual Xeon X5690 CPUs (up from dual X5650s). That's when the problems started to happen.

At first the system would boot up and run for about 2 to 5 minutes, then shut down and restart. Again it would run for 2 to 5 minutes and restart. Subsequent restarts would not even post before another shut down and reboot. I thought maybe the stock 875W power supply was not enough with the 2, 130W CPUs vs the old 95W CPUs and a 600W video card. I pulled the riser card and it would boot and run for awhile, but would randomly shut down and reboot. I replaced the power supply with a Dell 1000W unit. Still the same behavior. I thought heat maybe a problem so I installed software to customize air flow, no change. I finally narrowed it down to a bad RAM module. I replaced all 6 with like hynix RAM modules. Runs better, but still occasionally shuts down and restarts. Then installed a Dell combination SD card reader/DVD drive (this was just an upgrade with no thought that it would fix anything). System ran with occasional shut and reboot. Installed the riser card and got an amber power light and flashing 2, 3, 4 on the front panel. Removed the riser card and the system ran, again with the occasional shut down and reboot.

I bought a NEW riser card, installed the second X5690 and (with the recommended RAM configuration for a dual processor system) put it in the computer. Upon power up got a solid amber power light with flashing 2, 3, 4 on the diagnostic panel. Shut down and unplugged the combo drive (internal USB plug). Applied power and got a solid green light with 3, 4 on the front panel. Removed the riser card, plugged the combo unit back in and the system runs.

Riser card inspection showed no bent pins in the CPU socket, nor any bent pins where it connects to the mother board.

I am at a loss. Any advise would be greatly appreciated.

mazzinia_ · Answer

Put back the 2 original cpus , and test if the system is stable. Maybe the cpu installation had some issue / one "pin" or such is not making proper contact / the bios needs updating.

Start with reinstall the original cpus , that , as you yourself said , allowed you to do all the upgrades.

Test with all the upgrades, and start from there

amacmedic · Answer

@mazzinia_​ Thanks for your input. I will definitely do that sometime in January (2024) and post back the results.

amacmedic · Answer

Followup: I have tracked the issue of when the computer randomly restarting and I may have found the issue. The issue may be coming from the used video card that I purchased. I am thinking it may have bad RAM on the card because the issue happens more often then not when I am using an AI program for picture enhancement. It also happens other times during regular use, but most often when I am using the AI program.

mazzinia_ · Answer

An AI program would make use of the whole ram on the gpu, so it's a good suspect. Other suspect could be the gpu cooling solution... it may have been used with a waterblock and then the stock cooler reapplied but the thermal paste / pads over the ram etc are not ok.

Or the thermal paste was reapplied but not well and/or some of the parts that should be cooled too ( ram and power ) are not.

Or simply it needs to have new thermal paste and such.

amacmedic · Answer

@mazzinia_​ Thermal paste application was good. I removed the CPUs to check that. However I am now getting a RAM failure notification for slot DIMM 1 upon startup. Removing/replacing the RAM stick with a known good stick made no change. Cleaning the DIMM slot made no change either. Looking like a motherboard replacement may be in order.

amacmedic · Answer

UPDATE: 12-12-23

Did some research and found that the any update of the BIOS past A05 produced secondary processor function issues. Backdated the BIOS to A05 today. Installed the riser card and it runs fine with 2 processors.

For the past few weeks I have been getting a RAM failure notification for DIMM 1 when starting the computer. Removing/replacing the RAM stick with known good RAM did change anything. Cleaning the DIMM slot also produced no improvement. This may be where all my issues with random restarts may be coming from. Looks like replacing the motherboard may be in order.

Precision Fixed Workstations

Precision T5500 Green Power, solid 3, 4

Was this post helpful?