HP Z420 CPU Upgrade (E5-2687w v2)

[rtibby]

Hello all.

I upgraded recently a Z420 (left behind the Z400 to my kid as a gaming RIG).

I tested 2 CPU's in my Z420 : e5-1650v2 and e5-1680v2.

On custom air cooling , overclocked e5-1650v2 to 4.5 Ghz and the e5-1680v2 to 4.4 Ghz . With XTU set on all cores.

As a conclusion , the 6 core e5-1605v2 on all cores 100% is at 3.8 Ghz  / the 8 core 1680v2 is on all core 100% at 3.4 Ghz .

Basically the 4.5 / 4.4 Ghz is only on single core , or 2 cores active for 1680v2.

The 1680v2 passmark score is higher : 16351 (4.4Ghz) vs 14779 (4.5 Ghz).

Strangely I couldnt  go for higher speeds with 1680v2 compared with 1650v2 ... even though 1680v2 is rated at 85deg  , and 1650v2 at 70 deg. But , it does have 2 more cores ... and L3 cash is very good for 1680v2 : 25Mb compared to 12 Mb.

In general people say that the cooling solution is the cap for overclock speed. But what I experienced was that regardless of the temperature (which was low) there is a certain speed which if set , CPU instantly shuts down . In my case that was 4.6Ghz for 1650v2 and 4.5 Ghz for 1680v2 . I do not know exactly what is triggering this , because I've seen other users setting  higher speeds (for e.g Brian on this forum was stable with 1680v2 on 4,8 Ghz ...)

Anyway even though on all cores 1680v2 has a lower speed compared to 1650v2 I did keep the 1680v2 because of the overall better performance.

Z420 , e5-1680v2 @ 4.4Ghz all cores , Air CPU Noctua Cooler, 32Gb RAM HP 1xR8 8x4Gb , GTX 1070 , 1Tb SSD Samsung Evo 860, 1.5Tb HDD

[BambiBoomZ]

Rtibby,

You've done well to have an E5-1650 v2 stable at 4.5GHz. There are few 1650 v2 systems on Passmark at 4.5 and 4.6 and mostly on ASUS, MSI, Gigabyte, and etc. motherboards.

My secondary system, z420_3 with an E5-1650 v2 and using the special AIO z420 liquid cooler has never been stable at more than 4.3GHz.  I was fussing with 4.4 just last evening- unsuccessfully. This is the case with my E5-1680 v2 system, z620_2,   In both systems I don't wish to push the peak voltage over 1.5V as it deteriorates the memory controller over time. There is also a situation in which the all-important single-thread performance in most cases is better at lower voltages. On the E5-1680 v2,- which also uses a z420 liquid cooler, I still have had occasional XTU reboots, at 4.3GHz  with +217mV.  This occurs at odd time during booting.  I keep turning off items in startup configuration, but the problem is so very intermittent- once every four to six months, I can't certainly diagnose it. I've tried +238.75 and that also can have the problem.

What applied voltages are you using on your systems?

BambiBoomZ

HP z620_2 (2017) (R7) > Xeon E5-1680 v2 (8-core@ 4.3GHz) / z420 Liquid Cooling / 64GB DDR3-1866 ECC Reg / Quadro P2000 5GB  + GTX 1070 Ti 8GB / HP Z Turbo Drive M.2 256GB AHCI + Samsung 970 EVO M.2 NVMe 500GB + HGST 7K6000 4TB / Focusrite Scarlett 2i4 sound interface + 2X Mackie MR824 / 825W PSU /> HP OEM Windows 7 Prof.’l 64-bit > 2X Dell Ultrasharp U2715H (2560 X 1440)

[ Passmark Rating = 6280 / CPU rating = 17178 / 2D = 819 / 3D= 12629 / Mem = 3002 / Disk = 13751 / Single Thread Mark = 2368 [10.23.18]

HP z420_3: (2015) (R11) Xeon E5-1650 v2 (6C@ 4.3GHz) / z420 Liquid cooling / 32GB (HP/Samsung 4X 8GB DDR3-1866 ECC registered) / NVIDIA GeForce GTX 1060 6GB/ Samsung 860 EVO 500GB + HGST 4TB / ASUS Essence STX / Logitech z2300 2.1 / 600W PSU > Windows 7 Professional 64-bit (HP OEM ) > Samsung 40" 4K

[Passmark System Rating: = 5644 / CPU = 15293 / 2D = 847 / 3D = 10953 / Mem = 2997 Disk = 4858 /Single Thread Mark = 2384 [6.27.19]

[rtibby]

Hi BambiBoomZ. I used quite a few post of yours to learn about the HP systems...

So , I am using 242 mV on XTU for Z420 . I did use a reference a table on a post coming from Brian to estimate the additional voltage needed for the CPU.

About the cooling solution , as I said in my previous post although it is needed for avoiding high CPU temperatures , it seems it is not the only factor limiting the highest overclock speed. Still, for my Z420 I did setup a kind of custom cooling air solution which it seems works well ... https://h30434.www3.hp.com/t5/Business-PCs-Workstations-and-Point-of-Sale-Systems/Z420-Fan-control-s...

I keep temperatures low , and the original memory fan pointed to VRM and the 2 big nice Notua Fans of the CPU are cooling nicely the RAM sticks as well... And I did sorted out the Fans to speed up starting with lower CPU temperatures not like the original HP Fan management is doing ...

Anyway, I still cannot say clearly how the Bios determins the max overclock speed allowed to CPU ...

Z420 , e5-1680v2 @ 4.4Ghz all cores , Air CPU Noctua Cooler, 32Gb RAM HP 1xR8 8x4Gb , GTX 1070 , 1Tb SSD Samsung Evo 860, 1.5Tb HDD

[rtibby]

BambiBoomZ: "There is also a situation in which the all-important single-thread performance in most cases is better at lower voltages. On the E5-1680 v2,- which also uses a z420 liquid cooler, I still have had occasional XTU reboots, at 4.3GHz  with +217mV.  This occurs at odd time during booting.  I keep turning off items in startup configuration, but the problem is so very intermittent- once every four to six months, I can't certainly diagnose it. I've tried +238.75 and that also can have the problem"

I've experienced an interesting fact about rebootiong:  I was doing a Passmark test on the Z420 e5-1680v2 system : It went through all the test but it was rebooting at the end when was doing the Disk Performance...

I do have a 1Tb Samsung Evo SSD . And , Samsung Magician speed performance on . 

As soon as I deactivated Samsung Magician Speed Boost my sysyem was not booting anymore completing the test.

Now the reason of this behaviour is out of my understanding : why an overclocked system doesnt like the Samsung Magician Speed boost....

Of course after completing my test I activated again SSD speed boost which is a good thing in my opinion , giving a smoother experience using RAM.

Z420 , e5-1680v2 @ 4.4Ghz all cores , Air CPU Noctua Cooler, 32Gb RAM HP 1xR8 8x4Gb , GTX 1070 , 1Tb SSD Samsung Evo 860, 1.5Tb HDD