X48 Rampage Formula Preview.

[SoLoR]

Anyone have any idea how to stabilize ram? I have Q9450 and 4x 1gb Corsair Dominator CM2X1024-8500C5D (yeh i know 4 sticks is fail to begin with but...). I can run fully stable 8x450 with 333fsb strap and ram on 1080mhz and VERY tight timings on 2.0V, i managed to run inteliburn test with 8x462 and ram on 1110mhz 15 passes, however prime95 blend always fails in ~5min, no matter what i do, dimm voltage on 2.3V, common preformance level from 7 (using on 1080mhz)-10, clock twister changing also doesnt help (using stronger on 1080mhz, tryed moderate and lighter without change), dram static read on disabled (using enabled on 1080mhz). So if anyone have any idea what i can try more, for example that dram controler voltages? dram skews (was chainging them already, only thing i managed to do is preventing computer to boot up)? As soon as i droped memory lower my computer seems stable. Currently runing 8x470, with 400mhz fsb strap and memory on 940mhz and its prime95 stable for ~30 min already...

[mikeyakame]

Question. Are all the corsair sticks same revision? I came across a similar problem with 4 1gb corsair dominator sticks. Two were rev1.2 microns and would go beyond 1250mhz, and the other two were rev5.1 garbage they wouldn't even run more than 1080mhz.

It sounds like you have two different revision sets, Micron ICs need lots of voltage and you might find if you have a pair of rev1.x and rev4.1 or rev5.1 thats the problem. Rev4.1 and Rev5.1 don't like much more voltage than they are rated at. Where as Rev1.x Microns crave voltage like you can't throw enough at them.

Have a look at the sticks because it sounds like this is your problem in my experience. Also if Prime blend is failing you have Vtt or GTL related problems, nothing to do with memory. Different straps and dividers require different sets of values. 1:1 divider works great with 1gb sticks but runs like crap with 2gb sticks, and needs way too much Vnb. I had 4x1gb running on mine a while ago and managed to run them out to 1190mhz with some playing around so this isn't a problem at all.

RF runs 4x1gb fantastically.

[Brama]

RF runs 4x1gb fantastically.

Not mine

Probably due to a bad mix of sticks.

If you remember I asked you the settings used to reach 1100+ MHz, but I tried many times to reach at least 1000 MHz without any luck.

Probably if I could set skew individually for any stick, I can go upper than today but I am not so confident that a new BIOS release can add it.

It is the time to switch to core i7........

[screwtech02]

Not mine

Probably due to a bad mix of sticks.

If you remember I asked you the settings used to reach 1100+ MHz, but I tried many times to reach at least 1000 MHz without any luck.

Probably if I could set skew individually for any stick, I can go upper than today but I am not so confident that a new BIOS release can add it.

It is the time to switch to core i7........

Seems like it would be cheaper just to get some matched memory modules to me????

[mikeyakame]

Yeah bad mix of sticks is the worst of all. DDR3 fares much better with more sticks, DDR2 doesn't take that well to high frequency and 4 sticks no matter how good they are, the problem lies in its design. DDR3 design solved that problem.

I'll be going I7 soon too, 4GB ram isnt really enough for me to run full time Linux VM w/ 1gb ram! 4 2gb sticks isn't worth the hassle of restricting myself further, and spending more money on DDR2 yawn.

This has been a great board none the less, and I'm going to keep it around when I pick up I7 gear just because it's damn solid even when overclocked ridiculously

[Brama]

I confirm that my Adaptec 3405 is working with bios 0802. At last. Thx guys!

Finally they do it!
I am very happy for you. I remember how frustrated I felt when I had to use BIOS 403, as 407 didn't allow booting with my SAS disk controller

[Brama]

Seems like it would be cheaper just to get some matched memory modules to me????

You are right, but where is going the fun?

[mikeyakame]

You are right, but where is going the fun?

too right. Plus DDR2 is dead money nowadays, and no matter how much money you throw at a kit its still DDR2 and it's still at its design limits.

I'd still rather spend $500usd on a 6gb DDR3 1866 kit than spend 1/3 of that on DDR2 that works as it should.

[SoLoR]

Question. Are all the corsair sticks same revision? I came across a similar problem with 4 1gb corsair dominator sticks. Two were rev1.2 microns and would go beyond 1250mhz, and the other two were rev5.1 garbage they wouldn't even run more than 1080mhz.

It sounds like you have two different revision sets, Micron ICs need lots of voltage and you might find if you have a pair of rev1.x and rev4.1 or rev5.1 thats the problem. Rev4.1 and Rev5.1 don't like much more voltage than they are rated at. Where as Rev1.x Microns crave voltage like you can't throw enough at them.

Have a look at the sticks because it sounds like this is your problem in my experience. Also if Prime blend is failing you have Vtt or GTL related problems, nothing to do with memory. Different straps and dividers require different sets of values. 1:1 divider works great with 1gb sticks but runs like crap with 2gb sticks, and needs way too much Vnb. I had 4x1gb running on mine a while ago and managed to run them out to 1190mhz with some playing around so this isn't a problem at all.

RF runs 4x1gb fantastically.

they are same revision yes, however all are rev4.1. Also i kinda tryed CPU GTLs from 0.62x-0.065x and NB gtls 0.63x and 0.67x also VTT from 1.31-1.38 (real) and probably most of combinations and considering cpu voltage at 1.368V under load dont think its usefull to try any lower values and kinda want to stay under 1.4 VTT...

[Alien Grey]

they are same revision yes, however all are rev4.1. Also i kinda tryed CPU GTLs from 0.62x-0.065x and NB gtls 0.63x and 0.67x also VTT from 1.31-1.38 (real) and probably most of combinations and considering cpu voltage at 1.368V under load dont think its usefull to try any lower values and kinda want to stay under 1.4 VTT...

Try MemTest86+ V2.11 test. When it fails in test 5 it will be your memory.

My OCZ Reapers can run at 1120MHz with 1.92V. No matter what I try they don't clock a single MHz further without making errors.

[Michowski]

Ok so I want to set my gtl and skews now. I put myself at 430x8.5 and its stable with occt. Whats the first step in doing this? My real volts are,

Voltage Values
CPU Core 1.27 V
+3.3 V 3.17 V
+5 V 4.90 V
+12 V 11.98 V
+5 V Standby 4.80 V
FSB VTT 1.34 V
North Bridge Core 1.47 V
South Bridge Core 1.09 V
South Bridge PLL 1.55 V
DIMM 2.16 V
DIMM VTT 1.09 V

Current Values
CPU 20.32 A

Power Values
CPU 23.43 W

[mikeyakame]

Dial in 0.65x CPU, 0x63x NB, NB Skew Normal, CPU Skew -100PS, and go from there.

[mikeyakame]

they are same revision yes, however all are rev4.1. Also i kinda tryed CPU GTLs from 0.62x-0.065x and NB gtls 0.63x and 0.67x also VTT from 1.31-1.38 (real) and probably most of combinations and considering cpu voltage at 1.368V under load dont think its usefull to try any lower values and kinda want to stay under 1.4 VTT...

The problem is the rev4.1s, they aren't consistently good. Some overclock well, and others just poorly. Only good Dominator 1gb sticks are rev1.x, rest are pretty pathetic on Corsair's part.
Personally I would try swapping them between slots and see if it helps any, otherwise find the pair which clocks better and RMA the other and hope you get a better kit from RMA.

[Michowski]

Dial in 0.65x CPU, 0x63x NB, NB Skew Normal, CPU Skew -100PS, and go from there.

Ok il give that a shot when I get home. My question to you now is how did you take my settings i posted and determine which skew to try and how much gtl to use? I want to figure out how you took my specs to determine those four settings. Thanks

[mikeyakame]

Ok il give that a shot when I get home. My question to you now is how did you take my settings i posted and determine which skew to try and how much gtl to use? I want to figure out how you took my specs to determine those four settings. Thanks

I didn't they are just a good base set of values to start working with from your point and your cpu. You can't magically calculate what you need, you just need to start from a base working set, and make adjustments from there.

Plus I know from experience, that 45nm cpus seem to like 0.65x better, and 65nm cpus seem like to 0.67x better.

The NB & CPU clock skew values I gave you should be adequate for 400-500mhz FSB on this board, if you need to raise slightly then do so as a last resort.

[screwtech02]

Just ordered 8gb of this:

https://secure.xitrade.com/details.a...OCZ2FXE12004GK


Hopefully this and my "open box" Formula will play nice with each other...

[EOD]

Dial in 0.65x CPU, 0x63x NB, NB Skew Normal, CPU Skew -100PS, and go from there.

I've seen mention of the CPU Clock Skew, but nothing of the NB Clock Skew. Just for clarification, if you set the NB Clock Skew in addition to the CPU Clock Skew, will they me working against eachother? Is there any need for the NB Clock Skew to be adjusted?

[mikeyakame]

I havent found I needed to use any NB Skew except when I was playing around 1N command rate at DDR2-1066, 100ps NB, 200ps CPU made the most difference.

Otherwise CPU 100ps NB Normal should be pretty safe.

CPU being farther from the MCH/NB means it is more prone to clock drive deviation. FSB is driven right off BCLK so unless clock jitter directly created as a result of NB Voltage there shouldn't be much need to skew the NB clock unless it's for the purpose of aligning all the clocks with the IOH if its deskewed too much

There are 4 important Clock domains to consider.

CPU Clock, MCH clock, DRAM clock, IOH clock.

IOH (SB) has a few land pins which reference/sample directly from the CPU, so corrections to IOH Clock Drive skew are possibly made based on CPU Clock Drive skewing. I don't know for sure, but being so far away on the board it would make sense to reference the farthest Clock Domain to help out with strobe timing.

Host FSB BCLK is driven from the Analog PLL Clock Generator. This is located right below the Northbridge chip.

CPU Clock Drive is referenced from Host FSB BCLK.
MCH(NB) Clock Drive is referenced from Host FSB BCLK.
IOH Clock Drive is referenced from both Host FSB BCLK and CPU Clock.

DRAM Clock Drive is referenced from MCH Clock from what I can tell, since MCH is DRAM Controller and MCH is responsible for Read Delay timing during Cross Clocking Procedure.

Read Delay is the necessary turnaround between MCH request DRAM READ on one differential clock falling edge (ie low clock#), then a CPU requested FSB READ on the following differential clock falling edge (ie high clock). This is a procedure that has no direct handshaking between either end, rather relies on MCH reading data from DRAM, burst reading it onto the FSB host bus, then the CPU doing a blind read exactly half a clock period later whether data is there or not.



FSB BCLK sampling is done closest to NB, so NB Clock Skew will be the most resistant to deviation, and so far it seems from the responses of a few guys who are using same as me, CPU 100ps/NB 0s that this is very true. If I delay NB 100ps without delaying CPU 100ps more, and advancing DRAM CLK skew by 100-150ps, then I will get near instant BSOD's and pretty unstable operation even to then.

If your board needs it then thats ok, but unless you are using over 500mhz FSB and even more theres a good chance you probably should be safe with Normal NB Clock Skew.

100ps is a lot so unless the clock jitter is causing a deviation of more than half that amount (50ps), then don't worry. 100ps at 500mhz fsb is 1/20 clock period. It's quite a big amount and if its too much you'll end up more than likely corrupting CMOS or system files.

First thing I did was try different NB Clock Skew values on my board, and any more than 100ps and every time it resulted in no more CMOS data Had to redo all settings each POST, because they would completely vanish. NB Clock Skew is dangerous so believe me when I say if you don't need to change it then don't. It's only if you are pushing some heavy Vnb that it will show some serious benefits vs the risk of too much delay skew.

8GB Ram , 1.51-1.55v+ Vnb or Command Rate 1N + High DDR freq are the only cases I can think of which may need extra NB Clock Skew.

Edit:

Let me add one final concept to this post.

If you are highly experienced with fine tuning the bios settings and have a good understanding of the AGTL+ FSB design concept and how the CPU<->MCH<->DRAM procedure works at a low level, and you don't mind nuking a few OS installs in the process, then the following might interest you

With the precise balance of CPU Clock Skew, NB Clock Skew, PL phase pull-ins, GTL Refs, DRAM CTL Ref, DRAM secondary/tertiary timings and Vnb you can further tighten Performance Level beyond the point that is unpostable
Thats the key to tightening PL beyond the physical divider limit you reach in most conditions. Ie if PL=7 is limit for 12:10 divider at 500FSB, then PL=6 and even tighter is possible if you can balance the whole system on a pin, ie the timing of all strobes from end to end is near PERFECTION!

it's possible without a multi channel oscilloscope / logic analyzer, though you'd need a lot of patience,a good spare week or two to get it right, and a good OS ghost image to restore everytime you nuke it from being slightly wrong.

[Michowski]

Mike, I just tried your settings you mentioned before. I instantly failed intel burn test. With the skews and gtl on auto I passed it no problem. Perfect 5 out of 5. Manual it lasted .016 seconds before it failed it.

[mikeyakame]

Well that just means you need to Adjust them till it doesnt fail heh

Be patient you'll get used to it. Best way to make GTL adjustments is use Vtt as the coarse adjust value. With any given Vtt step cycle between 0.63-0.67x CPU and NB and see which values get you the farthest.

If its a higher ratio, then raise Vtt by one step, and drop back to 0.63x for both, and you'll find its pretty close to there. Same goes if if 0.63x gives you the longest run in Prime95, drop Vtt by one step, and start from highest multiplier and it shouldn't be too much the other way.

That's really all there is to adjusting GTLs. I can do it with my eyes closed because of the 100s of hours I've spent overclocking my board and I know exactly what to give it for which FSB frequencies and dividers, but I can't do that across the net for anybody. All I can do is give you a rough point to start with, and you shouldn't need to go too far from there.

Vtt must be set by the way if you want to change GTLs, if its on Auto you'll never it right. Same with Vnb and Vcc.

[Alien Grey]

FSB BCLK sampling is done closest to NB, so NB Clock Skew will be the most resistant to deviation, and so far it seems from the responses of a few guys who are using same as me, CPU 100ps/NB 0s that this is very true. If I delay NB 100ps without delaying CPU 100ps more, and advancing DRAM CLK skew by 100-150ps, then I will get near instant BSOD's and pretty unstable operation even to then.

That's true. The CPU and NB Clock Skew should be the same or CPU Clock Skew delayed 100ps more than NB Clock Skew.

[Alien Grey]

It looks like it's possible to use the ASUS PC Probe II version from the Rampage Extreme on our Rampage Formula board.

I tried it and everything looks alright. The only visible change I can see is the nominal voltage values.

[neo_rtr]

Hi, I friend mention this page and to be honest I am overwhelmed with the information that is given by you guys.
I looked around and I see that the Rampage Bios can be used on the Asus Maximus formula, which is the board I have. My problems since day one with this M/B is what no matter what I do I can not push it above 3.8Ghz using the QX6700 (did it few times but nothing really stable for long term use)
This thread is giving my hope in terms of doing just that.
Will have to flash my Maximus to Rampage first.

In the mean time.
Following the example of EOD post nr 1605 page 65 “I was running a Q6700 at 3.8GHz at 423 FSB x 9, “
My self was not able to boot. Went few steps up for the the volts in general with no luck.

On the QX6700 bios is 13.2 I have
CPU Ratio Control : Manual 8
- Ratio CMOS Setting :
FSB Frequency : 460
FSB Strap to North Bridge : 400
PCI-E Frequency: 100
CPU Voltage : 1,50625V
CPU PLL Voltage : 1.60V
North Bridge Voltage : 1,63V
DRAM Voltage : 2,02V
FSB Termination Voltage : 1,52v
South Bridge Voltage : 1,075v
Loadline Calibration : enable
CPU GTL Reference : auto

Can you give me a general advise on my existing setup before flashing the maximus to rampage.

Thank you

[EOD]

I actually found running too high VTT (FSB Termination Voltage) caused instability. I also found running the ram on a 5:6 (multiply FSB speed by 1.2) divider really helped with lower PL (thanks to XS). I also had great results running CPU GTL at x.63 and NB GTL at x.67.

I followed the advice here and CLR CMOS and started over with my OC. So I set the voltages, FSB, RAM timings only, and GTL while leaving everything else on auto, then went from there. It helped me lower my NB considerably for the same FSB OC.

[dizzy]

Nice to hear those good news about this 0802 BIOS.
But is someone so kind to send me the modded 0802 (with updated v8.5.1030 Intel RAID module, eventually also with updated JMicron368 module)?

PS: I didn't test yet 0802 BIOS, but being a BETA I would like to flash it with those updates included. I will post it soon and let you all know how it performs when got the file.
PS2: I have in raid0 2 WD750AYYS RE2 HDDs. This 8.5.1030 module BIOS will recognize the array or do I need to rebuild?


Thanks in advance...