did u test b4 and after for performance increases?
Printable View
a little something i grabbed before hitting a FSB Wall/Hole or even a voltage bug.
These is only at 1.65 volts in bios. Thing has some potential behind it.
http://img213.imageshack.us/img213/1927/4188ge1.png
Ill test for a much higher run once I water cool the chipset but this is about the best I can get for everyday use and not crash in my games...
it is what it is....I dont need that much vCore but didnt bother to change it...
Vista x64 and 4gigs of ram has been tricky for me so far...
Well it seems that its true Vista is a tad slower than XP but I have all the bells and whistles running with Vista so its not too much slower...
is there any mod for this motherboard???
After getting a chance to play with the Maximus this week I finally got it stable, huzzah!
The sticky wicket for me was getting the RAM stable. Two different brands in a 4x1 configuration wouldn't even boot in my old Asus P5W DH Deluxe at anything other that PC2-6400 speeds. I think that is a testament to the new X38 chipset.
Rest of my spec's in signature.
GAH! My board won't be here until Friday.......they just HAD to ship from the New Jersey warehouse rather than the Memphis one.....which is only 1 hour driving distance from my apartment. I guess they didn't have a part there....
is anybody running 1t on this board?
FYI - The stock ROG water cooling chart is up and rolling for the Maximus SE!
Come register your membership in the ROG SE club, and report your temps too. ;)
http://www.xtremesystems.org/forums/...d.php?t=164610
So far:
ROG Member ----- NB volts -- NB Loaded --------------- CPU ------------------
Renegade5399 ---- 1.65v ------ 34°C ------------ Intel Core2Duo E6850 G0 @ 4ghz
bill d ------------- 1.30v ------ 34°C ---------------------------- Q6600 Stock
ragge86 ---------- 1.65v ------ 36°C ---------------------------- Q6600 @ 3.8ghz
j0nnyr0773n ------ 1.54v ------ 44°C ---------------------------- Q6600 GO
Clump ------------ 1.62v ------ 46C (After 1HR or Prime 95) ------ Q6600 @ 3.6ghz
raju -------------- 1.65v ------ 49°C (Prime95 Torture Temp) ----- ??????
Note: After Raju remounted, max temp went down to 39°C
@MajoramQuote:
Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 9
FSB Frequency : 400
FSB Strap to North Bridge : Auto
PCI-E Frequency: 100
DRAM Frequency: DDR2-1066
DRAM Command Rate : 2T
DRAM Timing Control: Manual
CAS# Latency : 5
RAS# to CAS# Delay : 5
RAS# Precharge : 5
RAS# ActivateTime : 15
RAS# to RAS# Delay : 3
Row Refresh Cycle Time : 25
Write Recovery Time : 6
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: Auto
Ai Clock Twister : Auto
Transaction Booster : Auto
CPU Voltage : 1.41875 = 1.4
CPU PLL Voltage : 1.52 = 1.648
North Bridge Voltage : 1.49 = 1.520
DRAM Voltage : 2.12 = 2.224
FSB Termination Voltage : 1.40 = 1.472
South Bridge Voltage : 1.05 = 1.088
Loadline Calibration : Enabled
CPU GTL Reference : Auto
North Bridge GTL Reference : Auto
DDR2 Channel A REF Voltage : Auto
DDR2 Channel B REF Voltage : Auto
DDR2 Controller REF Voltage : Auto
SB 1.5V Voltage : 1.5 = 1.552
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 : 9
C1E Suppport : Disabled
CPU TM Function : Disabled
@Sdumperper
@Warman
Would you mind filling the table as above. It will be great as you're the only people getting your quad core at above 3.8GHz, many thanks for your help
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 9
FSB Frequency : 450
FSB Strap to North Bridge : 333mhz
PCI-E Frequency: 100
DRAM Frequency: DDR-1074
DRAM Command Rate : 2T
DRAM Timing Control: Auto
DRAM Static Read Control: Auto
Ai Clock Twister : Strong
Transaction Booster : Auto
CPU Voltage : 1.60000
CPU PLL Voltage : Auto
North Bridge Voltage : 1.80
DRAM Voltage : 2.25
FSB Termination Voltage : 1.86
South Bridge Voltage : 1.125
Loadline Calibration : Enabled
CPU GTL Reference : 0.67x
North Bridge GTL Reference : 0.67x
DDR2 Channel A REF Voltage : Auto
DDR2 Channel B REF Voltage : Auto
DDR2 Controller REF Voltage : Auto
SB 1.5V Voltage : 1.6
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 : 9
C1E Suppport : Disabled
CPU TM Function : Disabled
Hmmm, seeing as I have had no response to my last post on p63 i'll try to break it down a bit...
I have just updated my system slightly this week to include this Maximus Formula SE board, which is a 1.02g revision. Spec as follows:
Processor: C2D E6550 2.33ghz FSB1333 4mb
Memory: Kingston Hyper-X 2gb(2x1gb) 800mhz Low Latency 4-4-4-12 KHX6400D2LLK2/2G
Graphics: XFX 8800GTS 640mb
Sound: Auzentech X-Fi Prelude (in PCI slot above graphics card)
Case: Antec P182 SE
Power: Antec Neo HE 650w Blue
Drives: 2x 36gb Raptors, 1x 320mb Seagate 3gbSATA, 1x Sony SATA DVDRW
Now, its all setup and working. I have a Win XP Pro SP1 installation disk.
Despite using the Makedisk option(on another PC) so I could RAID0 my Raptors, XP setup was having none of it and kept rejecting the disk saying it couldn't create some sort of OEM file? In the end I have installed XP Pro on just one of the Raptors and I am using the other as the 'My Documents' disk. This is not ideally how I wanted it though of course.
Once I had everything installed and all the latest updates(SP2 etc) and antivirus installed I ran the 3dMark06 prog and got a score. Everything was fine and dandy.
Using the AI Suite I ran the auto system test button in the Gearbox area and then ran 3Dmark while using the Auto overclock option, which gave me about 400 extra on the score. I tried the Turbo option and this was slightly less than the Auto option. I then decided to start changing the multiplier in the manual overclock section. This was set at x7 so I simply changed it to x8 and rebooted. But it didn't reboot and the overclock failed. Fine, I thought I'd leave this manual overclocking alone for now and set about putting it on the Auto overclock option again. Now, every time I change it to that setting or re-run the system test for that option I get a BSOD. I did have Kingston KVR 667mhz memory in the system but I have just changed it yesterday to the Hyper-X 800mhz but it makes no difference, I still get the BSOD. Win32.sys seems to be the problem.
I have tried to clear the CMOS using the button at the back but I am not sure if that actually clears it? Have I corrupte my install and should I do a clean XP install, again? Have I broken something, namely my processor? If so I'll get it RMA'd asap?!?
I have now flashed to the latest official bios from the Asus webpage, 0602. Still same BSOD when changing the above settings.
Also, I am guessing its best to manually enter the timings for my memory in to the Bios as 4-4-4-12, as its showing up in grey as 5-5-5-15 on the Auto option??
Should I remove the black covers from the fusion block or not?
Lastly, every time I reboot my system since the above happened, the Wave volume is slid all the way to the bottom and the volume is Muted??
All help is greatly appreciated...
Great, gives an idea of what we could expect and what voltages we need for such a high OC.
On jmax-hardware ASUS interview about their X38 boards, ASUS representatives explained the PLL voltage usefullness. But they say that with a PLL voltage of 3.2v they noticed a degradation of their CPU overclocking capabilities.
As ASUS tends to overvolt in auto mode, I would check the PLL voltage for a manual setting one or two notches below the 3.2v limit
Thank you again for filling the table. Hope that Majoram and Sdumperper fill it also for a comparision
You got a great OC on air. But, I do believe more and more that the passive cooling of the motherboard won't handle such extreme OC sttiings in the short term when I see the NB voltage and FSB voltage you need (1.8v).
With your results, and if Majoram and Sdumperper confirm their settings, I'll make the step to the Maximus, remove the stock cooling, put two HR-09U on the mosfets with 2 silent 60 mm fans, my HR-05 IFX+80mm silent fan on the SB and cut th epipes from the NB fusion block to watercool the X38 chip, then give you my feedback ;)
Thanks again
warboy is that 450 x x9 prime or occt stable???
Hey jonny,
Wish I could help but it wasn't me that got that overclock above 3.8. It was just some person I chatted with in a computer store yesterday.
Sorry about that but I don't have any of his settings.
My system is still giving problems with 4 sticks of RAM even though I've tried setting the DRAM voltage to 2.3v
Maj
I can fill it out but I will wait until I water cool my chipset and get something more in line with my BlackPearl which I ran at 4.3 ghz for daily use. Also you wont be able to directly compare what I do with what others can do because I run my chips cool as can be seen in my sig...sorry I thought that it was known that Im a phase guy...
Also I cant go above 470 FSB right now even with phase...my chip has a known FSB wall of 490 so I know the chipset is holding me back right now. I am assuming that its just the heat so when I do add the water chipset cooling I should be able to push it up to a comfortable 480fsb or so...
Hi All,
I had an Intel XBX2 previously and it ruined two sets of RAM beause I used them in slots 1+3.. I'm using slot 2+4 now, even on my x38 mobo.
I've only got 2x1GB and I REALLY need another 2GB pronto (x64 Vista)
I want to get another 2x1GB but I'm scared that the mysterious memory problem will ruin them
Does anyone know the real reason behind this problem? or has anyone been running OCZ (particuarly the 1066 reapers) in 4x1GB config??
Many thanks
I just looked at your settings and have to verify that you are water cooling your chipset right? That NB and FSB termination voltage is very high. I keep my NB at 1.65 and FSB Termination a tat high at 1.58 but then again I havent started water cooling my chipset yet. What is your NB temp sitting at?
Thanks again for your suggestions.
Actually, my RAM is set to Auto. I haven't had time to tweak my RAM yet. I'm on vac for 9 days so I will have more time to tweak the settings and see if I can get more out of this board.
If I set the FSB Strap to auto are there any programs I can use to find which strap my board is using?
well increasing the NB and FSB termination volts to crazy settings helps but I dont like it on air so ill see what water can do for the chipset...
NB = 1.80
FSB Term 1.86 :eek:
My Max FSB for superpi at these settings appears to be around 476/477 and I was rock stable at 483 on my BlackPearl so I sure hope water helps the NB with the FSB max...
Nice OC on the Kentsfield Scott :up:
Finally going to cave in and yoink mine out of the DFI and see what this Maximus is all about.
Well first of all I would input all of the BIOS settings that your memory requires (voltage and timings), then I would start by set all of your settings to default, which is 333 FSB with 8x multi. Keep your CPU voltages at auto, and the NBV to ~ 1.4V or so.
If you still can't boot, then press F8 on boot and go into safe mode.
You don't EVER want to OC using any of these auto settings or in OS overclocking, because it will cause nothing but problems.
1T on Intel chipset is surely a no go, only Nvidia platforms could do 1T, that's based on all the boards I owned till now...
Explain to me how you are trying to set 8x multi on a 7 multi CPU, plz read up first all the overclocking guides regarding Conroe and then come back, you will kill something for sure if you just go trial and error...
You bought an enthousiast mobo that has insane voltages to select , also if you leave most settings on auto and push stuff it will overvolt all stuff... plz make sure before you OC that you are well aware what you are doing !!!
Damn multiquote doesn't seem to work on the laptop at work pfffffff
came across something weird, while running OCCT, the fsb voltage and dram ref volttage fluctuate to a really high value, over 1.8 each, this cant be too safe, also windows vista x64 stops responding but occt is still running, after i end process for vista not responding occt is still running stable, anyone else??
EDIT: the volts fluctuate only while using OCCT, prime blend is ok.
I occasionally enter bios to find FSB has been set to 2.0v, instead of the 1.3 I leave it at. If the setting wasn't RED, I probably wouldn't have noticed, bu I definately know I didn't set ever it that high...but it only seems to happen when I'm pushing the obard past the limit my other boards reach...
BTW make that 7x multi.....didn't realize your CPU used a 7x. You can only set higher mulitpliers on systems with unlocked multis. It's not the best way to OC even if it is available.
Basically you want to start off by making sure that your system is stable at stock before you start OCing.....i. e. making sure that the OS and driver installs are all correct.
I say memtest is not useful. I find generally 10mhz less than 24hr memtest, due to chipset. still wondering about that one....but...
prime small FFT for at least 12hrs
Blend for 24hrs
And you'll still get an occasional crash in IE, but only if chipset volts are a bit low.
OCCT does not load up cpu like prime. It also does a varied workload...some of like small fft, some like blend. but is only half-hour, unlesss you run custom test....so useless for me. I need to know the workload teh cpu is under while it crashes to accurately adjust settings in bios, and OCCT does not give this sort of feedback to user, like prime does, and if i want to monitor voltages, I'll stuff a grabber somewhere...
I still get in 0602 occasionally, but yes, not as often. Is almost rnadom, as I do not run much on this board ATM, prime95, cpu-z, coretemp, superPI, taskmanager...occt a few times....i thought maybe opening cpu-z and pcprobe at the same time caused it...but it's hard to catch it and start it at exact same momnet, when same processes are running in start-up...if i wait until fully idle, then nothing happens..
running on the 0402 bios, and my Q6600 stresses improperly. Core #4 stays at 100% after i stop the stress test after 2 hours while the other cores drop down to 1-2% load as they should. Core 4 will drop back down only after I restart the computer. Is my chip faulty?
Why did you have to RMA? How on earth did you change the multiplier in the first place? LAWL
Well my motherboard should be here Friday....can't wait to crack my knuckles on it. I have to install it along with a new PSU and my second 2900xt.
BTW guys if you're wanting to test your memory most accurately I use HCI Memtest. Prime95 blended is good all around test, but you can't use all of your memory. With this app you can actually specify the amount numerically. Touch your heatspreaders after running this for a while and they'll definitely feel warm to the touch.
http://hcidesign.com/memtest/
Everyone says the board overvolts. Has anyone checked this with a meter? If not does someone have a diagram of the leads and I could check them.
I could check them too......let me know where the leads are if you find out.
I've not read through the P5E thread, but are they getting the issues with Full Load and not loading 100%? Seeing as it's 995 the same board, I thought I'd question it?
http://img444.imageshack.us/img444/9...evdroophs4.jpg
EPU mod that was mentioned earlier?
So what does everyone think of 0701/0702?
I am still gonna stick with 0505 till they get some more things down.
I never get the not full loading but I think it's quad's only? Vdroop mod done and not done it's the same I think.
Moe than likely the same here, or even just go back to my Striker, at least it didn't have these funky issues...
I might try it over the next few days, but there is no guide as to how much to draw on it, what resistance we need to drop it down too... And I've killed about £600 worth of hardware before with a dodgy pencil mod (old ATi GFX card mod... went POOOOFF, took board/ram/psu with it... :( )
1.6v vCore, and everything else one notch into yellow I think? I'll check... Ram @ 400MHz divider 1:1 with FSB... 4-4-4-12, 2.2v... Static control thingy and the other ones all auto...
At 1.6v vCore I get the load issue, but with 1.55v (stable but not 24.7) i don't get it...
I'd care, yes. Look at the home forum of OCCT on hardware.fr (french only). I participated in the beta program. I was one of those who sputted on OCCT and its 30mn test. Now, in RAM mode, OCCT is for me the way to go.
OCCT new beta was completely redesigned for quad cores compatibility. My setup (in my sig) is stable Prime95 small/large/blend at 9x400 MHz. It never passed OCCT last beta in RAM mode above 390MHz. OCCT will stress the CPU less than Prime95, but the way calculations are made, stresses the CPU-NB-MB pathway a lot.
OCCT is not only a 30mn test. Most people banned it because of that 30mn message saying you're stable. Now, in new final release, it will be shown as: "no errors detected". The 30mn test is meant for a quick stability check only and can't be compared to a Prime95 12h run. You need to run a custom longer test, like with prime95, many hours to be safe
I won't go in a always hostile debate Prime95 or OCCT, but if you're OCCT unstable, than you're unstable. To rule it out, run OCCT in stock settings and you get your answer. You can ignore it, but you know, once in a while: you'll hang up...
I also agree saying memtest isn't a stress testing really, it only checks physical damage on memory modules. For unstable OC, it needs a really unstable one to fail
If OCCT would tell you what test it was running when it crashed, then it may be of use, but until it does, when it crashes, how do you know what to change? This is my only complaint about the program.
As I posted, I have run it, it IS part of my stability testing, but so is every other "Stress Test" Proggie with error-checking. It's not that it isn't useful...but it's not AS useful as Prime95 for me.
BTW, if you want to relate my comments to the OCCT writers, I'd much appreciate it.
"ASUS EPU intelligently switches between 4 and 8-phase modes, improving voltage regulator module (VRM) efficiency"
Snippet from Asus about EPU. Does anyone think our problems could be due to the switching modes on the fly?
what issues? max fsb no i dont think its related.
lol i stated this a long time ago. It's induced by the current draw of the cpu...draw goes up, 8-phase mode should kick in, due to EPU driver, but it isn't working, and we must await new driver or a bios that disables this functionality before the quadcore-load issue is fixed.
This is also why I mentiioned pencilmod may work, however it truly depends which is default mode...8-phase or 4-phase...and I can tell you based on behavior that it's not 8....but if it was, then i think we'd not have these issues, with droop-mod in place.
One good thing about it tho is that I won't be killing any more ASUS boards with quads....new P5K-pro(p35) has EPU as well...so may be a "new" standard feature from now on.
I'm still having stability problems with 4 sticks of RAM.
Other than raising the DRAM volts to 2.3v (is that high?), what else is there I can do?
Sorry I'm a newb at this, I have some knowledge but no experience.
My RAM is 4x1GB Crucial Ballistix PC2-8500.
I do a very simple test of playing the Crysis single player demo and it's always crashing.
I just changed over to 2x1GB OCZ Platinum XTC PC2-8500 and I just finished the whole demo with it only "freezing" (but not crashing) once. The game continued to play like normal.
Suggestions?
PS I'm tempted to just return the RAM and switch to G Skill 2x2GB PC2-8000 or buy another set of OCZ Platinum XTC (Platinum better than Ballistx?)
have you checked to see if your ram is running hot? make sure you have a fan over them.
What are your settings with 4 dimms? This is what im running for games now, mem not that high but 100% stable so far. 4x1024 PC8500:
http://img.photobucket.com/albums/v6...c/247-4096.jpg
Yes I checked and it's all outside the case and I put a fan sitting directly on top of them.
OS: XP 32 bit
(yes I know it won't recognise all the RAM but plan to move to 64bit Vista later)
CPU: Q6600 -> Not overclocked atm.
All settings are default unless listed below:
Strap: AUTO
Frequency: DDR2-1066
Dram timing settings: 5-5-5-15 (Subtimings not changed)
DRAM Static Read Control: DISABLED
Transaction Booster: DISABLED
FSB Frequency: 266
DRAM Volts: 2.30
With the OCZ RAM, I changed 2 things:
DRAM Static Read Control: ENABLED
DRAM Volts: 2.00
EDIT: Forgot BIOS ver 0602
Has anyone tested whether we get the same LOAD issue without Loadline Calibration enabled? Just increasing the vCore to compensate for the droop?
Oh, and 0702...not as good as 0701...just lost a 80% complete WU straight off the bat in F@H because it wasn't stable at the same settings! :(
doesn't work the same. at high speed, for me to get 1.55v, I would have to select 1.675v in bios, with droop figued in. do not want my cpu idling @ 1.675v!!!!
Yeah, I just tried it, exactly the same results. It did however stop the "System" resource hogging the CPU load...it only stole between 7-10% instead of a solid 25%! So maybe this is what is meant by the vDroop mod helping? Disable Load Line, vDroop mod so we don't have to set crazy idle volts and then it get better? No idea here, just throwing out thoughts?
Yes, it works that way. I am unsure of the current limit built into circuit, but the closer you get to that limit, the larger the "drop" in utilization. Likewise, the farther away from the limit, the less drop in utilization, and hence my comment about droop mod being better, as the circuit would be partially unaware of the actual current drawn by cpu. It's like the driver tries to "latch" the extra 8-phases, but it cuases a loop that does not ocmplete due to utilization of the cpu by prime. If we triggered this change when idle, i think there would not be a problem...the process jsut is not getting enough cpu time.
But that in itself causes issues...
I think I'll just stick with 3.825GHz for now, as that is OCCT/Prime/Folding/3D2K6 stable and has no issues at all.
Idealy I'd like it with the 8x mulitplier but 9x425 is ok for now...
Out of interest, what volts and settings are people using for 475MHz+ on the FSB/NB? I seem to need about 1.9v to get it anywhere near remotely stable... :(
So what's the final verdict on the black rubber caps on the Fusion block?
If you use air cooling only, should you take them off or leave them? Or perhaps it simply makes no difference either way.
I took them off, I didn't like the idea of hot air building up inside and being trapped there... Northbridge temps dropped maybe 1 or 2 max! Nothing major... TBH, I think reseating with AS5 would do more good but I've run it so much that I'll probably have no chance of removing the heatsink now! :(
Leave them on. If they are off, dust will build up inside.
OK, so now i am having a hard time getting OCCT ram test stable, i have ran prime for hours with no problem, small and blend, but occt fails after 30 mins, memtest passes as well, and voltages fluctuate when running occt, not prime, could someone with the same set of ram run OCCT 1.1.1B on custom and run the ram test for 1 hour, would really appreciate it, also windows vista x64 is my OS
I have no issues with OCCt being stable. maybe up chipset volts a bit, if ram is known stable on other board, otherwise, look at upping ram .05v.
Running 4GB pc D9GMH @ 1120mhz 5-5-5-18. TRAS of 15 WAS NOT STABLE AT ANY SPEED FOR ME, over 1040mhz.
nope, vista32 and XP32. Might be an issue with X64, but i am not aware of such an issue with OCCT. Try TRAS change as well...I realized a bit back that TRAS on DDR3 is much higher than DDR2, and this chipset was intended for DDR3, right?....
http://img152.imageshack.us/img152/8108/8xcpule1.th.png
Just by clicking on the Apply button... They make it too easy to make mistakes!
Well, I have RMA'd my processor because its now causing the BSOD everytime I try to do anything other than standard? Surely I must have done something to it? Or have I just corrupted my XP install??
I'm having a lot of issues where USB hubs (such as those on a monitor or a card reader) are losing devices (well actually just not finding devices), when this happens I need to reboot to get it to be detected. I'm not sure the cause, as these same devices were never lost on my old MB using the same layout. This leads me to believe that this is a MB issue, but I wanted to ask it anyone else has had a similar issue.
Also while the USB devices will work if detected on resume from hibernate, most often one or more are not detected.
ERIC
OK guys, i have to say, in the beginning i loved this board, but as of late, im beginning to hate it, it might be specifically mine, but im having bunch of problems that are very annoying, mainly my specific to my ram, just wont be stable, even at rated speeds, anyway i have packaged my board and am going to return it while i still can to my local fry's..........guess my participation from this thread will probably be gone but good luck to everyone else, im going to wait it out for the 780i to SLi my 8800GT..... :welcome: