Friends of 0 and 1,
I probably have made a silly mistake.
In upgrading a z620 that arrived with one CPU and DDR3 1600 ECC unbuffered RAM, I added 4X 8GB of DDR3 1600 unbuffered RAM for a total of 40GB. After adding a pair of Xeon E5-2690's plus updating the BIOS and chipset /firmware:
HP z620 (Rev 2) 2X Xeon E5-2690 (8-core @ 2.9 / 3.8GHz) / 40GB DDR3-1600 ECC) / Quadro K2200 (4GB) / HP Z Turbo Drive (256GB) / 800W > Windows 7 Professional 64-bit > HP 2711x (27" 1980 X 1080)
[ Passmark System Rating= 5322 / CPU= 19675 / 2D= 767 / 3D = 3544/ Mem =2337 / Disk = 12951 ] 8.15.16
The CPU number was very odd in that the overall system rating of 5322 was the highest of the 177, z620's tested on Passmark and yet none of the individual scores was the highest. The Passmark CPU rating of 19675 was below the overall average of 20699, the 21257 of the top E5-2690 rating in a z620 and up to 22731 in a z820.
For a couple of days, I fussed with CPU-reltaed settings in BIOS and minimizing baxkground processs and etc, but I had the thought that the other dual CPU systems I've had used ECC registered which has to do with creating a one cycle wait as a buffer to synchronize the second CPU.
Today, checking the z620 manual, it specifies that 4GB modules may be unbuffered but 8GB RAM modules should be registered.
This is more of a puzzle than a dramatic effect, but might the unbuffered 8GB modules be having an effect on dual CPU efficiency?
Cheers,
BambiBoomZ
HP z420> Xeon E5-1660 v2 / 32GB DDR3 1866 /Quadro K4200 (4GB) / Samsung SM951 256GB + Intel 730 480GB + WD Black 1TB > Win 7 Prof'l