UnOfficial Asus Rampage II Extreme Thread

[cstkl1]

well hmm i also had a psu related mem timing issue

silverstone op1000/da1200 i could run memtest hci with trFC 48.. thats he screenies i showed some pages back
plugged in hx1000 tRFC now is 60.
plugged back da1200 and its 48. no idea except the hx1000 rails are not as strong as op1000 and having pc probe show the 3v,5v,12v for hx1000 results in monitor blinking every few seconds.
each 285's now running on different rails just to make sure.

btw on the nb solution
think its just the thermal paste under it
took out the whole thing and used tuniq tx2
take note tuniq tx2 application method is different
u need to put quite a bit and it has to be viscous and not dry for it to work.
now on passive cooling in casing temp on memtest hci load is 55C only even better when it was in the bench rig.
hr05 solution was around 48C on load
also now my vtt voltage for uncore 4000 dropped to set 1.5375, read from probe it 1.48v

[dejanh]

well hmm i also had a psu related mem timing issue

silverstone op1000/da1200 i could run memtest hci with trFC 48.. thats he screenies i showed some pages back
plugged in hx1000 tRFC now is 60.
plugged back da1200 and its 48. no idea except the hx1000 rails are not as strong as op1000 and having pc probe show the 3v,5v,12v for hx1000 results in monitor blinking every few seconds.
each 285's now running on different rails just to make sure.

btw on the nb solution
think its just the thermal paste under it
took out the whole thing and used tuniq tx2
take note tuniq tx2 application method is different
u need to put quite a bit and it has to be viscous and not dry for it to work.
now on passive cooling in casing temp on memtest hci load is 55C only even better when it was in the bench rig.
hr05 solution was around 48C on load
also now my vtt voltage for uncore 4000 dropped to set 1.5375, read from probe it 1.48v

Ha! Interesting observation. I have been doing some testing over the last few days, and one of the tests I did just last night is I put the memory timings back to 8-8-8-20-1T from 7-8-7-18-1T when running 4.2GHz and 1600MHz memory. I booted perfectly this morning. Tonight, I will set the timings to 7-8-7-20-1T and see if it boots tomorrow morning, and then I will set them back to 7-8-7-18-1T and see if it boots fine.

Overall, though, this is very strange because it seems like it could be the PSU not being able to give enough power to the DRAM to start up at certain timings, which to be honest I have never seen or heard of before. Once the system boots, I can run memtest86+ and memtest hci for days on end at 7-8-7-18-1T and never get a single error. I can keep rebooting the system, shutting it down, booting up again, as long as I do not leave it off for prolonged periods of time. Extremely strange.

Anyway, I will update more as I run my next set of tests as specified above over the next couple of days.

Maybe I ought to pick up a Silverstore Zeus or something and test this out as well...

[dejanh]

Ha! Interesting observation. I have been doing some testing over the last few days, and one of the tests I did just last night is I put the memory timings back to 8-8-8-20-1T from 7-8-7-18-1T when running 4.2GHz and 1600MHz memory. I booted perfectly this morning. Tonight, I will set the timings to 7-8-7-20-1T and see if it boots tomorrow morning, and then I will set them back to 7-8-7-18-1T and see if it boots fine.

Overall, though, this is very strange because it seems like it could be the PSU not being able to give enough power to the DRAM to start up at certain timings, which to be honest I have never seen or heard of before. Once the system boots, I can run memtest86+ and memtest hci for days on end at 7-8-7-18-1T and never get a single error. I can keep rebooting the system, shutting it down, booting up again, as long as I do not leave it off for prolonged periods of time. Extremely strange.

Anyway, I will update more as I run my next set of tests as specified above over the next couple of days.

Maybe I ought to pick up a Silverstore Zeus or something and test this out as well...

Well, last night I changed the timings to 7-8-7-20-1T and this morning I could not boot. I will restore the timings today to 8-8-8-20-1T and will report back tomorrow on what are the results of the change. I am anticipating that I will boot successfully as was the case yesterday.

[dejanh]

Well, last night I changed the timings to 7-8-7-20-1T and this morning I could not boot. I will restore the timings today to 8-8-8-20-1T and will report back tomorrow on what are the results of the change. I am anticipating that I will boot successfully as was the case yesterday.

Well, as I have anticipated, I booted successfully this morning. Timings were 8-8-8-20-1T.

So, now I know this for my Dominators set at 1600MHz...

If I have timings at 7-8-7-18-1T I am 100% Memtest86+ and Memtest HCI stable once I boot, but a cold boot in the morning will always fail.

If I have timings at 8-8-8-20-1T I am also 100% Memtest86+ and Memtest HCI stable, and a cold boot in the morning will always succeed.

Now why in the heck would this be the case? Could it really be the PSU not being able to power up the ram at lower timings? Why would this occur? I have never seen behavior like this before.