The Maximus Formula Thread - Help/OCing/Guide/Rampage conversion

[Grnfinger]

grn, better get that 4870x2 under water soon, I'm coming after it


I always find the deals and hard to find hardware

Geebus thats a nice price

Still waiting for stock...no stock soon and I'll order one from USA

[Jodiuh]

Wow? Those stores are still open, I suppose I better get down there.

[Zucker2k]

grn, better get that 4870x2 under water soon, I'm coming after it

Chew, Grn, you want a piece of me? Look in my sig. Broke 27k with my x2 flashed to the Asus Top bios @ 832/1000.

[devilhood]

I would stay away from Crucials and this board, I've had about 3 pairs of the ballistix 8500's and they never worked correctly on this board. After switching to G.skill have not had a single issue for almost over a year.

I can confirm the same also... Crucial gave me some really bad experiences with this board.

G.Skill = the good stuff

[mikeyakame]

DDR2 Channel A REF Voltage :
DDR2 Channel B REF Voltage :
DDR2 Controller REF Voltage :

Can I get some instie on how and when to tweak these settings and for what to where? I had them left to Auto for over a year and recently fixed to DDR2 REfand I only noticed it moved my DDR ref from 1.152 to 1.130. at some settings it would also auto to 1.20. Where do I want this to be concidering my types of high FSB OC and 5:6 divider?

Also, I have moved to the OCZ FLEX EX PC2-9600 and it the mobo appears to need some more tweaking to keep it stable at 1200MHz/2.256v in 5:6 at 500FSB. outside of lowering the trfc to 55, what can I do about the rest of the sub timings or learn more about them? Most notably, the secondary settings if any benifit is gained. My Primary settings are 6-6-6-3-55-6-3 Auto was 6-6-6-3-46-6-3.

DRAM reference voltages are only really useful above 1250mhz or if you are running 2.35V+

DRAM CTL Ref works best with negative values, I use -20mV or -30mV in extreme cases, it helps to manage clock jitter at high Vdimm.

DRAM CHA/CHB Ref no matter what I set them to give inconsistent results period. I leave them on Auto, thats the only way they work correctly. Every bios I've used has been the same, they just work correctly on Auto and thats it.

your timings should be.

5-5-5-15-3-55-6-4 8-3-5-4-6-4-6 14-5-1-6-6

secondaries can be changed to 9-4-5-4-6-4-6 if stability is an issue. If you change tRTW to 9, lower tWR by 1 since both those timings are write recovery turnarounds.

Secondary timings are for consecutive bursts thats why there is very little to gain on 2GB sticks, there are too many banks which means longer Refresh Cycle turnarounds, longer Write Recovery turnarounds, etc for same rank if the burst (read followed by write, write followed by write, read followed by read, write followed by read, write followed by precharge, read followed by precharge, and so on) requires access to data from more than 4 banks in same rank. tFAW (four-bank activation window) is the timing responsible for this, and it is not adjustable.

Try that as a starting point for 1200mhz.
The most gains from secondary come from changing tRTW (8) lower if you can get it to run error free that is. You can always compensate by doing, tRTW = 7, tWR=7 you will gain a little bandwidth that way.

You'll find 2GB sticks don't have much room for adjustment on secondary and tertiary timings, they work best near default. 1GB sticks respond much better with tight secondary/tertiary timings.

[chew*]

Chew, Grn, you want a piece of me? Look in my sig. Broke 27k with my x2 flashed to the Asus Top bios @ 832/1000.

I'll give ya a go, probably going for a DI run this weekend, May need LN2 to top ya though. I might have to dig deep in to my pocket of tricks to catch you on DI. FYI they do 700/1200 maxxed out ccc already.21,7K @4.0g not optimixed no tweaks, air

[Zucker2k]

I'll give ya a go, probably going for a DI run this weekend, May need LN2 to top ya though. I might have to dig deep in to my pocket of tricks to catch you on DI. FYI they do 700/1200 maxxed out ccc already.21,7K @4.0g not optimixed no tweaks, air

To be fair, you'll have to show some of those tweaks You already have an advantage by going extreme cooling. I'm only on water.

[Grnfinger]

Chew, Grn, you want a piece of me? Look in my sig. Broke 27k with my x2 flashed to the Asus Top bios @ 832/1000.

I can do 24500 with NO oc on the gpu, cant overclock the GPU yet cause of the lack of proper cooling.
Its -30c here today so I have extreme cooling if I open the window
Theres a single 4870 at work that might migrate to my house for a little tri fire run. Getting pissed about the lack of proper cooling.
My new case, cpu block and rads are here, just need the ing gpu block to get things up and going. Want to start my i7 build shortly if the tax man is not to hard on me.

[chew*]

To be fair, you'll have to show some of those tweaks You already have an advantage by going extreme cooling. I'm only on water.

Allrighty anything goes I may need to harness the power of hybrid with my 4870 and 1 4850

[loften]

after playing with the e8500 at 420x9.5 for a few days i wanted to see if it would do anything.. posted windows at 465x9.5 cpu 1.26/1.23 real and fsbt 1.30/1.22 real.. it wouldn't go untill i upped those volts a little from the 4.0 settings but only a little and it posted windows no problem.. since the air cooling i have is pretty much maxed out at 4.0 and the and idle temps went up from 38 to 41 degrees i shut it back down and went back to 420x9.5 untill i get water cooling for it.. once i do i think it will be on like donkey kong..

[JLK03F150]

I could use a little advice on pushing this farther. It's having some stability problems at 4.2GHz. It can run small ffts for a while, but blend fails in less than 5 minutes. Safe to assume I got something wrong with the ram or mobo settings. Hopefully someone can spot something I am missing.
E8600 E0 @ 1.262v, apogee block, 100% load temps ~42-47C
Maximus Formula SE 1207 bios
G.Skill PC2-6400 HZ (D9GMH) 4-4-4-12 @ 2.0v Worked great on my old Abit AW9D at 420-430MHz.
Seasonic S12-650W
2x HD3870
120GB sata 3.0 HDD, CD-RW, DVD-ROM

Extreme Tweaker 
Ai Overclock Tuner : Manual 
OC From CPU Level Up : AUTO 
CPU Ratio Control : Manual 
- Ratio CMOS Setting : 10
FSB Frequency : 421
FSB Strap to North Bridge : 400
PCI-E Frequency: 100
DRAM Frequency: DDR2- 842
DRAM Command Rate : T2
DRAM Timing Control: Manual 
CAS# Latency : 4
RAS# to CAS# Delay : 4
RAS# Precharge : 4
RAS# ActivateTime : 12
RAS# to RAS# Delay : 3
Row Refresh Cycle Time : 40
Write Recovery Time : 8
Read to Precharge Time : 3

Read to Write Delay (S/D) : 8
Write to Read Delay (S) : 3
Write to Read Delay (D) : 5
Read to Read Delay (S) : 4
Read to Read Delay (D) : 6
Write to Write Delay (S) : 4
Write to Write Delay (D) : 6
DRAM Static Read Control: A
Ai Clock Twister : A
Transaction Booster : A

CPU Voltage : 1.2620
CPU PLL Voltage : 1.5
North Bridge Voltage : 1.31
DRAM Voltage : 2.00
FSB Termination Voltage : 1.40
South Bridge Voltage : 1.075
Loadline Calibration : E
CPU GTL Reference : A
North Bridge GTL Reference : A
DDR2 Channel A REF Voltage : A
DDR2 Channel B REF Voltage : A
DDR2 Controller REF Voltage : A
SB 1.5V Voltage : 1.5

NB LED Selection : NB Volt 
SB LED Selection : SB Volt 
CPU LED Selection : CPU VOlt 
Voltiminder LED : 

CPU Spread Spectrum : Disabled 
PCIE Spread Spectrum : Disabled 

Advanced CPU Configuration 
CPU Ratio Control : Manual 
- Ratio CMOS Setting : 
C1E Suppport : Disabled 
CPU TM Function : Disabled 
Vanderpool Technology : Disabled 
Execute Disable Bit : Enabled 
Max CPUID Value Limit : Disabled 


USB Configuration 
USB Functions: Enabled 
Legacy USB Support : Enabled

[loften]

i would disable loadline calibration set the cpu volts at auto and fsbt volts at auto nb volts at auto and restart go right back into bios then see what they say.. work from there.. keep cpu and fsbt volts closer together.. 1.40 fsbt is quite a bit at that clock especially with the cpu volts being that much lower... i would say right from the get go that the nb volts need to be 1.47-1.51.. to get anything above stock speeds stable. my guess is that you should be able to hit that clock with both cpu and fsbt closer to 1.26-1.28 cpu and 1.28-1.30 fsbt. another thing is my 2x2 corsair 4,4,4,12 6400 gave me errors at anything above their rated speed but they liked the looser timings on this board on the 400 strap 5,5,5,18, tested them at 1066-1081-1119 speed with the recomended volts and dram static read enabled, ai clock strong, transaction booster disabled (0) performance level 7.. passed hours of tests at those settings but would fail at 4,4,4,12 at right around rated speeds even going up or down on their recomended volts a little. (EDIT) here are my settings for mine at 9.5x420 100% stable on air with loadline disabled.. vcore 1.26875-1.232 real.. cpu pll 1.50-1.623 real.. north bridge 1.51-1.552 real.. fsbt 1.30-1.232 real.. south bridge 1.050-1.088 real... dram 2.08-2.208 real,, 4x1 crucial ballistix double sided in it now 5-5-5-18-3-50-6-3-8-3-5-4-6-4-6 @ 1119 on the 400 strap.. again dram static read enabled, ai clock strong, transaction booster disabled (0) performance level 7

[JLK03F150]

I had been stable at 4.1 with slightly lower vNB & vcore. I raised the vNB to 1.41v, lowered FSBt to 1.28v & it seems better.

Trying 4.3ghz (10x431) w/ 1.28vcore, 1.30v FSBt, 1.45vNB.

[Kiwi76]

loften, I'm not understanding everything you say.

My Northbridge is perfectly stable at ~1.4V (1.39V set in BIOS), with a dual core and two sticks of RAM. Since his 2GB has less address space for the MCH to handle than my 4GB, there's no way he needs at least 1.47-1.51V for stability. This is using a 400MHz strap, but I think a 333MHz strap works fine too. My RAM isn't so good, so I use a 400MHz strap and lighter timings and performance levels for that stuff.

Also, my E8600 is running on a Maximus Formula at 4.5GHz with the FSB Termination at only .05V above the Intel recommendation of 1.15V. It's set to 1.3V, which results in ~1.2V. It does 4.0GHz at stock FSB termination. It may even do 4.5GHz at stock FSB Termination. I don't know. I haven't tried it.

I was under the assumption that FSB termination and CPU PLL is really only important for quad cores, at least not on dual cores until really high FSB speeds.

I do have Loadline Calibration enabled. It lets me get away with less voltage at idle, and voltages seem to stay the same under load and at idle (it'll go up from 1.376V to 1.384V every here and there, but it stays under 1.4V I have set in BIOS for it never to go over).

Am I missing something? I don't understand the every detail of every setting, but my values are way lower than you're suggesting, and it works.

[loften]

i am new to the 45 nm so maybe i shouldn't be giving any advise especially if it may be wrong... you guys got it good with your boards to run such low volts on your nb... mine won't on either sets of ram i have, 4x1 or 2x2. i will say with loadline enabled i can get away with lower volts on both cpu and fsbt. right about where yours is. v droop, i read a couple articles about loadline enabled on 45nm causing issues not under load at idle or coming down from load.. i do know his fsbt was way up there more then normal for a 45 nm.. with loadline disabled my real volts are only a little higher then they were set before with it enabeled even though they are set a little higher in the bios settings now there is only a little difference between the way they are @ real volts. my temps didn't change and i feel much safer especially reading about loadline being enabled may cause problems with 45nm..

[Grnfinger]

loften, I'm not understanding everything you say.

My Northbridge is perfectly stable at ~1.4V (1.39V set in BIOS), with a dual core and two sticks of RAM. Since his 2GB has less address space for the MCH to handle than my 4GB, there's no way he needs at least 1.47-1.51V for stability. This is using a 400MHz strap, but I think a 333MHz strap works fine too. My RAM isn't so good, so I use a 400MHz strap and lighter timings and performance levels for that stuff.

Also, my E8600 is running on a Maximus Formula at 4.5GHz with the FSB Termination at only .05V above the Intel recommendation of 1.15V. It's set to 1.3V, which results in ~1.2V. It does 4.0GHz at stock FSB termination. It may even do 4.5GHz at stock FSB Termination. I don't know. I haven't tried it.

I was under the assumption that FSB termination and CPU PLL is really only important for quad cores, at least not on dual cores until really high FSB speeds.

I do have Loadline Calibration enabled. It lets me get away with less voltage at idle, and voltages seem to stay the same under load and at idle (it'll go up from 1.376V to 1.384V every here and there, but it stays under 1.4V I have set in BIOS for it never to go over).

Am I missing something? I don't understand the every detail of every setting, but my values are way lower than you're suggesting, and it works.

Different straps, PL level and trans booster will effect the MCH and how much volts it will require for stability.
Alot of ppl here will tell you high NB volts, but its is not required for some boards, mine only needed 1.45 for 489FSB 333 strap PL7 with ram running 1175MHz.
Alot of guys will over volt the hell out of everyting and call it a day, I preferr to find the lowest stable setting.

[JLK03F150]

Trying 4.3ghz (10x431) w/ 1.28vcore, 1.30v FSBt, 1.45vNB.

Sadly, that wasn't stable. I'll get back to testing this weekend though.

[ShawnTRD]

Hows the new BETA BIOS 0802???

[chew*]

No clue haven't tried it but if its like most of the newer bios its not for me.

I barely have time to play with this rig anymore.

[neo_rtr]

I have managed to flash my Maximus to Rampage 0802. I am using a QX6700 and the 1st thing i notice is that i can run at much lower voltages.

http://valid.canardpc.com/show_oc.php?id=508970






If any one can tell me how to get to 3.8Ghz i will be extremly happy..
I have tried lots of things with out luck.

[.OCX]

Continued crashing ? Not sure why?? I tried upping the voltage dropping the voltage same w/ NB/SB/ RAM even everything set to auto continued to be a problem!
Went back to Old Gigabyte GA-X-38-DQ6 motherboard it has been running Prime95 stable 14hrs+ as I have obtained 3.8 only thing that bothers me about this board is its vcore jumps around under cpuz even with loadline cal enabled?? I have no idea why yet the asus board was rock solid voltage no matter where I set it? I might invest in a DFI board soon if this gives me any probs? Thus far it seems stable as a pyramid @ 3.8ghz ran 4 ghz for about an hour but couldn't find stable setting so dropped to 3.8 and hey Im ok with that for now, though DFI might be the answer?? Not sure why the asus max formula kept freezing real shame as I like the board and it bios immensely

viccan

http://www.anandtech.com/cpuchipsets...spx?i=3184&p=6

Loadline Calibration may be crippling your stability.

Imagine our confusion as we desperately struggle to understand why our system is Prime95 stable for days yet continues to crash under absolutely no load.

I have a QX9650 arriving from newegg tomorrow so I'll have more too contribute on Maximus Formula + QX9650 findings in a couple of days.

[Zucker2k]

http://www.anandtech.com/cpuchipsets...spx?i=3184&p=6

Loadline Calibration may be crippling your stability.



I have a QX9650 arriving from newegg tomorrow so I'll have more too contribute on Maximus Formula + QX9650 findings in a couple of days.

Please tell me newegg sent you that chip for free. If not, send it back and get a Q9650. You can use the balance for a whole host of other components, including a Rampage Formula x48. Just a friendly advice.

[.OCX]

Please tell me newegg sent you that chip for free. If not, send it back and get a Q9650. You can use the balance for a whole host of other components, including a Rampage Formula x48. Just a friendly advice.

It's a Q9650 lol sorry about the $700 dollar typo, I even had to double check my order.

[MUYA]

It's a Q9650 lol sorry about the $700 dollar typo, I even had to double check my order.

Please do keep us informed when you get your Q9650.

I recently got one and with the mobo on 1302 bios (not rampage modded) I can't get anything stable over 3.6 GHz in windows. The CPU is watercooled with D-tek fuzion and MCR-220 rad. TH north bridge is also indcluded in the loop.

Also, on any cpu temp monitoring tool, I get core 1 that is higher than all the other cores. Is this normal? I even took the block off and re-clean, re seated, re-pasted AS5 and re installed the block on but still getting on core higher. Happens idling or loaded

For e.g. it would be like ;
core 1 = 42 Deg C
core 2 = 35 Deg C
core 3 = 37 Deg C
core 4 = 35 Deg C

[Grnfinger]

Please do keep us informed when you get your Q9650.

I recently got one and with the mobo on 1302 bios (not rampage modded) I can't get anything stable over 3.6 GHz in windows. The CPU is watercooled with D-tek fuzion and MCR-220 rad. TH north bridge is also indcluded in the loop.

Also, on any cpu temp monitoring tool, I get core 1 that is higher than all the other cores. Is this normal? I even took the block off and re-clean, re seated, re-pasted AS5 and re installed the block on but still getting on core higher. Happens idling or loaded

For e.g. it would be like ;
core 1 = 42 Deg C
core 2 = 35 Deg C
core 3 = 37 Deg C
core 4 = 35 Deg C

X38 chipset is 45nm quad limited