Memory Tuning – Part 1 – DDR4 MFR – My first fail.

Fail1As majority of serious Overclockers know, not every day and attempt goes to plan.

My plan was to start a memory tuning series of articles next week with DDR4 MFR and working forward of the IC’s that have been released.  It was not to be the ultimate how to tune guide but more of a sharing of my trials and tribulations.

After getting the first OC<$100 done and dusted, I pulled out the GIGABYTE Z170 SOC FORCE and set a basic 4.5Ghz overclock.  Perfect for efficiency testing.

I started by testing each stick at stock settings with everything set on Auto.  No worries, got all 4 sticks through XTU no worries.  Then I went and set XMP and tried again, no worries all sticks passed.

Net step was to set a few timings and freq and bin the sticks based on the lowest stable volts.  I have played with this sticks before on X99 and know they can do it.  So the process begins with 3200 12-15-15-28, trfc 274 and the command rate to 1T.  Then setting DRAM voltage to 1.60V and working my way up in 0.01V increments until it got through XTU.

Stick 1 would not boot at all, giving me Code 55 everytime.  So this one is a dud.

Stick 2 would boot from 1.62V and get into windows.  Walking it up to 1.75V it still wouldn’t pass XTU.

Stick 3 would boot from 1.61V and get into windows. No stability all the way up to 1.75V in XTU.

Stick 4 exactly the same as stick 3.

I felt a little defeated, but was not to let this beat me.  I had set the SA and IO CPU voltages to 1.35V on the firs tlot of runs.  So I went back through with those 2 volts set to Auto, then 1.1 and 1.2.  Still nothing to show, so I took a break for the night.

I went back today and continued on from where I left off and set SA and IO Voltages at 1.3 and then 1.4.  Still nothing.

 

It is then after another 4 hours, I decided enough was enough.  I could not for the life of me get these MFR sticks stable.

I honestly feel a little disappointed that I have nothing to show for the hours I put in but that’s Overclocking.  It’s more upsetting that this was to be the stat of a new series for Extremebenching.com.   I will now move onto Samsung E Die IC’s and try again.  Not having any D Die it is the next logical step.  So thanks for reading and over the next couple of months I should hopefully have some results together for E Die, AFR and B Die.  Then we can move backwards to DDR3.

 EDIT – This is by no means any fault of the motherboard or memory, this is a just a very finnicky IC and will more than likely end up being my settings.  I have not given up just yet though as I know it’s possible on this motherboard, I just need to figure out how.

Thanks for reading, JD. Also a massive shout out to Leeghoofd and Sofos for their help so far in trying to get this running.