Try this settings for 2815~3050 CL12, CR1T
Lat. comp enabled
Dram swizzling bit 2 disabled
Dram init. value 59
Dram rtl chan. A = 55/55/55/55
Dram rtl chan. B = 56/56/56/56
Dram I/O chan. A = 9/9/9/9
Dram I/O chan. B = 8/8/8/8
Try this settings for 2815~3050 CL12, CR1T
Lat. comp enabled
Dram swizzling bit 2 disabled
Dram init. value 59
Dram rtl chan. A = 55/55/55/55
Dram rtl chan. B = 56/56/56/56
Dram I/O chan. A = 9/9/9/9
Dram I/O chan. B = 8/8/8/8
Oddly enough, those are the exact auto-set RTL and IO-L values I'm getting with my 4x8GB MFR kit at least. Does CR1 improve stability on this board vs CR2? I saw that mentioned in one of the guides but it seems to defy the conventional logic that setting the command rate looser would increase stability when overclocking. I've also read advice to disable Swizzling Bit 2, I wish I knew exactly what that is doing.
The reboot loops are the most annoying thing, you don't even get a solid LED code that you can use to start your troubleshooting process.
After taking out 2 DIMMs, switching to the black slots, much fiddling with BCLK lower than 125 but on the same strap, finding one that boots, and using initial BCLK settings, I finally managed to get the thing to POST and boot Windows with the TridentX at 3000C12, but it fails 32M instantly. And the RTL settings look wonky to me.
http://i.imgur.com/rA0doJM.jpg
Well, since I got this far, I figured I'd try my luck sticking the other two DIMMs back in. To my surprise, it booted into Windows. I had enabled Additional Training so I'm not sure if that played into this or not. First attempt at 32M failed, so I upped VDIMM from 1.75v to 1.85v and re-ran...and it passed. This seriously took me the whole evening. I had best disable Swizzling Bit 2, key in these RTLs manually and save a profile because I don't want to go through all that again.
http://i.imgur.com/9I2nkT1.jpg
read this my friend
http://i.imgur.com/hnMDfWE.png
Unfortunately, trying to manually enter the RTL values displayed and saving didn't work and now I can't boot back at the same OC :mad:
Why does that work?! Man this board is quirky!
^It is best to test each slot with each module and find the slot that said module goes highest in, it's just Haswell being Haswell:p:
shiranu i did this for you
GSkill TridentX 2800CL11-13-13-35 2T Hynix CFR@ 3000CL11-14-13-35 2T 1,825 (just used the same profile of the TeamX didnt change anything yet)
http://i.imgur.com/GOgzDL2.png
and now i took pictures for you
http://i.imgur.com/OsBVIHD.jpg
http://i.imgur.com/dMqNXuU.png
http://i.imgur.com/tneyrVy.png
http://i.imgur.com/4vTLMb7.png
http://i.imgur.com/nRKvov6.png
http://i.imgur.com/qQl4gj4.png
http://i.imgur.com/eQtyMyN.png
http://i.imgur.com/S5iuXfi.png
tRFC 500 good grief that's loose...
No go I'm afraid with your settings. Reboot loop as usual.
CFR CL9 tights probly max on air. 2.02V
http://img849.imageshack.us/img849/7951/siy5.png
http://img138.imageshack.us/img138/9526/78wz.png
Why are you guys wasting so much time to stabilise 3000Mhz ram clocks with amasing crap performance ?
tRDRD and tWRWR at 6-7 is such a waste :shakes:
Why climb a mountain? Because it's there.
Well let me know if they fix this board within the next 20 days or so within my store return window, I'm going back to my "it just works" OC Force.
^^Imo, the board is ok. at least with my memory clocking. CFR (in general) is a bit difficult.
If other board works then probably you should stay with it.