Try:
10 X 380
333/1333
MCH up to 1.425
CPU @ 1.35V
That should get you to 3.8 with the memory @ 1526. Leave the memory sub settings auto and run the memory @ 7-7-7-20 at your 1.82V.
Printable View
Also try 60,4,15,7,7
I was trying my damnedest to hold out for a BT2 deal, bit I bit on this, to hold me over until my Intel hookup can wrangle a Q9450. Credit card companies love me.
Down the road, I may move to a Maximus2, or a BT2, but I'm gonna start playing now.
Now to go pickup some DDR3....
-bZj
Well I tried those sub timings and at:
Fsb 450
Ref Freq 400
Mem Freq 1333
Mem Timings 7-7-7-20 60,4,15,7,7 2T
It loads all the way up to just logging in and then it just goes blank, I tried changing the timings from 7-7-7-20 to 9-9-9-24 and it actually let me log in but froze up shortly after..
I'm running at 400fsb now and it runs OK, any other sub timing suggestions? I would really like tio get it running at the same clocks as i did successfully in Bios 1478. Also i must mention that I'm running 4 stick of DDR and not just 2..
RC,
Does your memory have an XMP profile? If it does, then under the Performance Memory Profiles option on the memory page you should see Automatic, Manual, Profile1, Profile2, etc.
This might be useful, if you have XMP memory. If you select automatic, press enter, then select manual press enter, it will load the default SPD settings for all the timings and sub timings into the manual setting. If you select Profile 1, press enter, then select manual it will load all the XMP profile timings and sub timings into the manual setting.
You sig says 4GB DDR3-1600, is that 4 x 1GB sticks, or 2 x 2GB sticks?
The Corsair 1600 XMP that I have has loads 88, 5, 10, 6, 6 for it's subtimings when selecting Profile 1.
The OCZ 1600 XMP that I have loads 72, 5, 10, 6, 6 for it's Profile 1.
One thing to keep in mind, for tRFC, the first number in the subtimings list a larger number is faster/more aggressive. All the rest lower is faster/more aggressive.
From the timings you last posted 60, 4, 15, 7, 7 you might try raising tRRD up to 5 or 6.
i doubt this would make the difference, but i run my vFSB at 1.35 instead of 1.325. i doubt that's your problem though.
Thanks for the tips Anubi, yes my memory does have the Intel XMP profile (profile 1 in bios) and I have tried that as well, unfortunately it uses the same timings and sub timings i have tried previously. The info regarding the first sub timing (higher number=faster timing) I was un aware of, this could be very helpful as I was always increasing that number to try and get stable..
P.S. I'm running 4x1G sticks of DDR3-1600
I guess what I'm missing is some sort of "proceedure" to set the sub timings to a workable setting, is it recommendable to change multiple sub timings at once or is it better to change just one at a time?
RC,
Just read your last few posts again. Are you sure it's memory frequency that's causing your trouble? A good way to verify would be to set the memory frequency to 800 (really slow) and see if you still have problems.
anyone got vcore for this mobo ?
CPU Core voltage can be set up to 1.9v. Is that what you wanted to know?
nop, vcore, 1.9V in bios, is 1.84V real and that is way to low
Well, I'm not 100% sure the problem is the memory, I just assumed so because I was unable to get to my standard overclock (with Bios 1478) in any bios newer than that. One of the reasons I thought that was the addition of sub timings in the bios.
I'm still just struggling along to get to a stable overclock in any bios newer than 1478, which now that I think about it its not absolutely neccessary, I can just stick with the older Bios unless I'm bored and want to "fiddle".
I'll let you guys know if I'm ever successful.
Has anyone tested the 1554 yet? I see no comments on what changed. I've seen the log, know what's new but wanted to know how it feels, any better than 1521?
I guess tRD/Performance Level still not there, right? Hardware monitor still the same?
I wonder if anyone ever get to use the power button on the board, the little switch that's in the board. I have occasionally but what I really wanted to know was if anyone, those fearless ones, have tried soldering a similar switch to the existing but unused reset switch next to it, and if it works. I believe the board is prepared to have it installed, can't understand why Intel wouldn't bother to put it there, would be helpfull.
Well i figured out the problem, the voltage isnt changing when I set it. I verfied this with everest (my old board displayed correct voltage), also when i set voltage offset it does work but it runs way to hot as the voltage is too high. Is it time to RMA my board?
It looks like GJW had this problem as well, did you ever get it resolved.
Windows update has mentioned that there is a driver update for Vista Audio drivers from IDT. However, the drivers available in Intel website are still the same. I guess they didn't have the time to get them there still.
I was reading an article regarding the new Gigabyte boards to be released with the P45 chipset and it seems they are also jumping on the UEFI (aka The Framework) bandwagon, as an alternative to the good ol' BIOS.
Why is it that Intel, which was the initiator of this long awaited (at least by me) BIOS killer, is the only one not working on it, or, at least, talking about it? They have started it and then spawned it as the Tiano project, but no news or improvements have been known of.
I would like to see it for our BT, I keep saying this with hope that the Intel guys are reading this forum.
Yes I had that problem. I received a new board from Intel and it went all bad. So I bought an Asus Maximus Extreme and moved forward with that. Eventually Intel did offer to exchange the DX38BT I had for a DX48BT2, which I accepted. Last week I put together another machine with that board and a QX9770 CPU. I am still running the 1521 bios version. It has the exact same problem, can't change CPU voltage. I get back home tomorrow and I will load in the new bios and see what I can do. I also could not get the beta version of IDCC to work (running Vista Ultimate 64 bit) but there is also a newer version of that. If that works, I'll try to change the CPU voltage through that as well. I'd rather do it in the bios, but if that's the only way that works.....
I'll report back when I have answers.
j
So far I have had this problem on the one DX38BT that I had functional for a while and now I have it on the DX48BT with all totally different hardware. Different CPU, different RAM, different everything.
I do have another DX48BT that I will build up in the next few months, but I'm going into it thinking that I won't be able to change CPU voltage.
j
My DX38BT should be arriv9ing any minute now.... Sweet!
Time to break out the Rocketfish & HX620, and order some H2O kit.
Now just waiting on Aeneon DDR3 to arrive....
-bZj
Latest bios still no control over voltage. I'm glad my primary machine is an Asus board.....
j
Tony at OCZ did a good write-up about tuning DDR3 at 1600+
http://www.ocztechnologyforum.com/fo...ad.php?t=38357
Nobody's talking about the new 1554 BIOS. So, what's it like? Is it worth the update? Can anyone oc any higher with this one? At least it is possible to go as far as 720MHz for the PSB I believe, right? Well, that's way behond normal proc capabilities.
Anyone tried also IDCC v4.1.0.168?
the 1554 BIOS is like sunshine in a bottle. everyone with a Bone Trail should have it installed. i haven't tried the new IDCC yet though. my previous experiences with it were so bad that i'm waiting for the green light from you guys before i install it again :D
MB
I was waiting for comments before jumping in.
Any changes in the mem voltage granularity? I believe a 0.1V was already in order.
Also, the PCIe freq with a 1V increase/decrease would make sense.
Hardware monitor is also a little on the poor side, but there are also limitations to the monitoring chip, the datasheet shows this.
I guess tRD or performance Level (as Asus calls it) are still not there.
I have 1554 on both the BT and BT2 with no new issues. Stable before, stable now.
Can't stand IDCC so I can't comment on it (except for the "can't stand" part). :)
Thanks you guys, guess I'll have to try it.
C1E and Watchdog on the regular BIOS menus is reason enough, I think.
Also, given that new things ARE showing up, someone must be listening.
Submit a ticket for tRD. I did.
Did you email them directly to the support email?
I'm not sure they even read it through, their first answer is "Intel does not support overclocking", although thay state otherwise in their page.
Support and marketing guys there must hate each other...
Yep, Just say "BIOS Feature Request, add tRD memory control".
Tier 1 is still clueless but BIOS seems to be a keyword to get it the correct group.
Thanks gtj, I'll do that.
I'm off to bed now, tomorrow is an early rise for work.
night
:clap::clap:tony great job thanks!!!!!!!!!!!
My proccessor will still not overclock (at any voltage) anywhere beyond 3.4ghz with any bios newer than 1478 even at stock fsb speeds.. I have spent countless hours trying to figure it out.
My RAM is going out tomorrow. I should have a BoneTrail to match my BadAxe by the end of the week!
-bZj
Quick little price update:
DX38BT + Ghost Recon Advanced Warfighter 2 = $149.99
http://www.compusa.com/applications/...CCODE=COMEM054
Just bit, and purchased myself a board. Hopefully it will be as enjoyable as my XBX2 with the latest bios!
Yeah Tony, great job indeed!
Can you hint on some of the new stuff coming up?
New audio drivers came up on Intel site, can't see any changes really on the readme file, there's even a mistake with the driver version in this file. It seems it must be for Vista SP1 only, since they say you need to uninstall and re-install after SP1, minor update I guess.
in regards to the x48 motherboard which is mine now for a couple of days. i've
wondered if i should be concerned with cooling the power regulators (mosfets)?? when using a tower cpu cooler such as the thermalright ultra 120? also, another question, in the intel spec sheet they indicate thermal sensors on the cpu, nb and sb dies, i assume readable somehow or as something other than what i've assumed in other boards, ie.,xbx2. what i'm getting at is within everest with the typical cpu,gpu,hdd and such there is also a motherboard temp and an auxillary temp indicated. now, i love the program know as everest but i not taking thier word as what a sensor should be called. could it be, for my own curiousity, that cpu, mainboard and auxillary
be temps for the cpu, nb and sb. anyone want to enlighten me?
I wouldn't worry about the mosfets. They run pretty cool.
The 3 temps are...
CPU (the diode)
Motherboard (the aSC7621 monitoring chip itself which is between the cpu socket and the first dimm socket)
DIMM (the sensor is in the space between the second and third dimm slots)
ok, thank you gtj, i'll buy that. funny thing is when reading to many message
boards is that i care about stuff i never cared about before.:)
by the way, for the modest overclocker like myself, i'm as happy with the x48
as i've been with xbx2 the past year and half.
Okay all you guru's....tried first with old version bios, then most recent...
I am not a total rookie overclocking... (this PC is a e6400 clocked to 3.2 on BX2)...but never worked with this chipset
googled to hell looking for resolution.
I am trying to overclock my CPU (Q6700) on this x38bt chipset. I have 1333 DDR3 corsair RAM...here is my issue.
I tried a 10x and a 9x and I cannot seem to overclock CPU stable to (even) 3.4 GZH, without enabling "cpu voltage offset". '
I have up volts for fsb and mch... ect.
I can set volts all the way up to 1.6 and no dice, and YET with a voltage of 2.875 and offset enabled (+300mv = 5.875v which is less than 1.6!) I am stable....
I should be able to overclock this CPU without enabling offset right?
by the way temps are okay as I am using TEC cooler
P.S. if someone has numbers for q6700 (or 6600) overclocked perhaps I can try them...
baffled
You should NOT need the offset.
Start with 333 * 10. You should be able to do that at stock voltages. My q6600 G0 did it with no problems.
You should not need to touch vMCH or vFSB until you close in on 400FSB.
What are your memory settings?
running loose timings of 9 9 9 24...1066 @ 1.7v as per man specs...
333x10
all default volts...
all eist watchdog ect disabled...
no post. Time to consider possibility of bum chipset/ cpu?
hmmm 1.78 allows me to post, (and up) take to BSD when I try to let windows boot up
Hey craftsman
It's funny (well, not really), but gtj is right, those are the temps available for this board. Still, according to Intel both the MCH and ICH should provide temp readings as well since both have internal diodes. However, we don't get to see those in hardware monitoring. The limitation I believe is in the aSC7621 which has no more diode inputs, one internal and 2 external. However, if Intel did it right I believe we could see them through SMBus I guess, in addition to those. Not sure why Everest can't see them though. This chip is very limited regarding available readings.
It's not really a limitation of the asc7621 but more a limit on Intel's part. Yes, both the X38/48 and the ICH9 have diodes but neither exposes them to the outside either via pinouts or registers. The ICH9 is also on the SMBus (it's the controller) but it's slave capabilities are limited to reading watchdog and rtc info. The ICH9 is also on the PECI bus with the cpu and asc7621 but the ICH9 and asc7621 are reading from the cpu. The ICH9 can't be read that way.
Manel, i've considered giving the intel integrators toolkit a whirl because that seems like a possible way in. sorry though, i'm yea, nah,yea, nah, about it. you might like to try though, eh.:)
GTJ, does your BT2 use the same chipset drivers as the BT does?
gtj, you really dig into this deep man. Yes, the IC9 is the SMBus controller and I was guessing that because of that there could actually be a way of accessing those readings. If you say that's not possible I'll buy that, you know what you're talking about for sure. And I thought maybe Everest could get it that way also but I guess not. But following that line of thought, what are those diodes there for? If they are not accessible from the outside and if their readings are not available they are of no use. Maybe just for internal control but what will happen if temp reaches higher limit, will the MCH/ICH shutdown? What about the rest of the system, will it be forced to shutdown as well? How will it know if the readings are not passed through? Well, too many questions here!!! PECI would indeed be another way but guess not too.
craftsman, I might try it. I have also been like go/no go with it, just to have a feel, see what additional potencial it can bring, if any. Let you know if I try it.
xMrBunglex, since the MCH is exactly the same, only renamed really and better binned, the drivers/INF should be the same, don't know why Intel hasn't updated them to reflect the "new" chipset though. It was time Intel would clarify the issue with the version of the INF files provided in their site and the CD that comes with the board, though. Never knew which are the latest and which should be installed.
Still, X58 has already been leaked, triple DDR3 channel controller is now integrated on the Nehalem CPU so no more MCH designation. QPI will replace the current interconnection finally, AMD had the advantage here with HT. Funny how Intel is always trailing after AMD nowadays, as much as it's hard to admit, but still wearing the performance crown. Also option for 2xPCIe x16 or 4xPCIe x8 seems nice. Too bad no SLI is still in the horizon. All this in Q4'08, near enough.
I started reading hardware manuals for fun in 1970. :D
Anyway, the diodes are there for the internal shutdown logic and in the ICH9's case, it's actually a basic environment sensor and fan controller for lower end motherboards that don't have separate sensor chips.
The cpu, mch and ich can all signal an overheat which would cause the system to shut down.
Guest if your still reading this, you didn't mention if your changing the FSB clock. You mentioned your memory is 1066 rated for 1.7v, that's not very good actually. Any reaonsable quality memory rated for 1066 should be able to do it at 1.5v, if your not changing the FSB clock.
woohoo my $150 Bonetrail came in today.
Now I just need to transition from my BX2 and get some other hardware parts. I'm going to pick up an E8300 for $160 @ Ewiz. Do you think 470FSB is obtainable on this board with a dual core with the latest bios?
Another BadAxe2 convert. :)
I'd liketo join that club, if I could get hardware that worked. I bought used DDR3 (Aeneon XTune 1333), and it won't POST. I get the 3-beep salute, no matter the slot or DIMM count.
Seller is working with me, so hope I'll get it going soon.
Letting hardware rots sucks.
-bZj
Hello,
I would like to buy one of the Intel DX38BT - Motherboard.
Does anyone know, if MDT DDR3-Ram is compatible with this board?
Thank you
down8
There should be no issues with the Aeneon I guess, but if you can't POST you may be experiencing my initial nightmare with this board.
I had such a problem with my OCZ Plats but it was due to an early BIOS version which was very picky with the memory. Do you know which BIOS ver you have in this board? You need a stick of mem that works, in my case it was an Asus 1333 stick - go figure, didn't even know Asus had mem. Then flash to the latest BIOS and you're set.
You need to try some mem, OCZ has a reputation of being tough to boot, but maybe Corsair, Crucial, you need to try.
It took me over a month to get it to boot, just to find the right mem.
Good luck.
Stilo
Try Intel site for the compatibility chart, I believe MDT is not there. Try and get someone to try it on their mobo, it's the best way. You never know with this board. Latest BIOSes should be ok.
I'm up!
The seller sent some Crucial to test with, and it's installing Vista now. I had some initial problems with video. Had to switch to another card, then boot and setup BIOS, then replace my HD3850.
Can't wait to play. I should have an X38 waterblock today, and a CPU block this coming week, too!
-bZj
Hey down8
Great to know you're steaming away already.
Let us know how it plays.
IDCC won't start. EDIT: It starts now - sometimes.
RAM wont run over 1066. XMP shows in 1554, but when selected board drops back to 1066 JDEC.
EDIT: IDCC can change the RAM & CPU, but I can't change (reliably) from the BIOS.
EDIT2: CPU-Z reports 667MHz for RAM, IDCC reports 1600MHz. Which to believe?
Re-reading this whole thread for tools/tips....
-bZj
I installed Everest 4.5, but it won't show me the CPUID - pay only function?
When I try to run it, or the mem benchmark, I get "Invalid floating point operation!" and an empty window.
-bZj
CPU-Z also incorrectly reports my memory as DDR 1333 when it's actually at DDR 1480. all other monitoring programs read it correctly. for the XMP profile, do you have your memory in slots 2 and 4? those are the slots you need to have your RAM in to use the XMP profile. and it sounds like your system isn't stable yet. it took most of us a while to get our systems running the way we wanted and i don't think there are any beginners in this group.
Attachment 79981
Attachment 79984
Attachment 79982
Attachment 79983
I have been happy to find that it is possible to run the XMP profile #1 with
4 x 1gb of Crucial Ballistix PC3-12800. With a FSB of 425 this puts the
memory at 1700mhz.
I have recently switched processors from a Q6600 to a E8500. The E8500
needs 1.35 volts to operate at 4ghz. While massaging the various parameters
in a effort to lower the core voltage, I decided to try a lower multiplier.
(500 x 8). I made the adjustment in the bios, leaving the memory on profile #1. The Bone presented me with a nice blue screen with instructions to run "disk clean" AND to reauthorize Vista! I had to do this, why I don't know
except some corruption must have happened to the ms files.
Later on I gave it another try, but this time I lowered the memory to a 1:1
ratio before changing the cpu. When it booted up it stopped before windows
with a notice that a file was missing or corrupted and that I needed to run
Vista startup repair. I did this and was able to boot into Vista this time.
Operation was as expected and passed all stability tests. However, it wasn't any better than the previous 425 x 9.5 setting.
As others have noted, this motherboard does not like changes! Almost any memory change will result in a "no boot" condition. Incidently, it has not been possible to make any changes at all in the gtl settings. One click off the automatic number = "no boot".
After the Bone has been adjusted to the settings that "it" likes, it is perfectly stable in all respects. However, it should not be such a hassle to get to this stage. This bios is by far the worst I have ever encounted, you would expect better from Intel!
Yeah, I was trying to goof with the GTL settings as well and came to the same conclusion.
Also, the "other settings" for memory don't work (tRFC, etc). You can set them to anything you want and they seem to be ignored.
your memory benchmarks are low. my memory is running a lot slower than yours and my Everest benchmarks are a lot higher. timings vs. MHz?
http://i21.photobucket.com/albums/b2...1/membench.jpg
Attachment 80002
Yeah, this is the best I can do with four dimms. I can tighten up the timing if I remove two. I got some improvement above by raising Vmch which is working harder with four modules. Also, maybe you have better memory with the new Micron chips? I have the D9GTR ICs.
What other programs should I try, to check the RAM speed? Everest won't work, CPU-Z is unreliable, and IDCC is questionable.
-bZj
Down, use SiSandra. the free version of their program has a memory bandwidth and latency benchmark. expect your Sandra scores to be lower than our Everest scores though. i get about 9000 MB/s on the Sandra memory bandwidth test compared to 9800 on Everest.
Sarge, i don't know why kind of chips my Patriots use and forgot the website that listed all of them. if anyone wants to link me, i'll give you a quarter. i don't think my RAM has the best IC's though. i can score over 10,500 when i clock my memory higher, but i found better overall real world performance with slower memory and tighter timings. that's just based on my own testing though.
MrBunglex......just to see if the NB was bogging down because of the 4 sticks in my system, I pulled two Dimms and set the memory to your exact readings.
My numbers are still much lower than yours even with a higher cpu speed. I actually was surprised that my Crucial would run at 6-6-6-16! Incidently, all the sub-settings are the same as yours as read on memset.
The one unknown is what memory slots you are using. I am in the blue slots which is correct according to the manual. I have read somewhere that the black slots give better performance, but I wouldn't know why since Intel specifies that the first slot must be populated.....any comments?
This is interesting in that it shows four Dimms apparently do not load the north bridge if you raise the voltage. Maybe I will try the black slots later.
Attachment 80027
MrBunglex........I gave the black slots a go with no improvement. I am giving up. Back to 4 Dimms. At least I found that my Crucial Ballistix would run at the tighter timings, but not as good as yours. Who knows why???
no clue. i'll put up all my BIOS settings sometime in the next few days so you can compare yours to mine if you want. i do use the black DIMM slots for my memory, but i think it's only necessary if you want to automatically use the XMP profiles.
My XMP profile showed while using the blue slots.
Neither showed to work, however. The blue actually was closer to 'working' XMP.
Hello everyone,
Quote:
My XMP profile showed while using the blue slots.
Neither showed to work, however. The blue actually was closer to 'working' XMP.
Well I have the same Crucial Ballistix 2 x 1GB as you, and of course neither works here.
I have the DX48BT2 tho, and the processor is the QX9770. I got everything matched for a FSB at 1600 mhz at stock. (My idea was to get a 4 GB kit of memory, but since I saw that the memory is not there yet - is just this gamble with this profiles and stuff I only got this until the "true" JEDEC DDR3 1600 is released).
Ok in my case the computer boots right away as long as I don't load any of the profiles; whenever I load any of the 2, I got in the infamous boot loop (yes watchdog is off), but every time it managed to boot with the profile even if it takes it a while.
I tried with more than one bios with the same exact result.
I ran memtest without any errors with the first profile I imagine would do too with the other, the modules seem to be error free.
I contacted Crucial and was a waste of time, the person said that they didn't release yet XMP parts, (.....?!) and to just load the basic timings and voltage and set the advanced timings in auto... well too bad there isn't auto for the advanced timings, neither I think would do any difference.
How many of you are affected by the boot loop problem, and what modules are you using? This Crucial works right away here if NOT profile is loaded, otherwise might boot at first try but most likely will not if is a cold boot.
I halted to move the system there until I find a way to solve this, since I don't like the stress this issue puts over the system.
Please any settings to make this work at stock settings for me (That is 1600 mhz FSB at 1:1 ratio) and alternatively if there is any memory that really works? :(
Thank you for your help.
So, you got it to boot @ 1600 w/o touching anything?
I had IDCC show me 1600 once, though CPU-Z showed 1333.
Yes, memtest confirmed it. It is running at 1600 mhz.
But it takes a while for it to boot. That is what I am trying to solve.
It boots with any of the profiles from 1.9 (as it read in the profile) to much lower 1.74 ... tried all that to see if I could solve it. I left it at 1.78 V.
Which version of memtest shows you the RAM speed?
I still can't get it to boot @ 1600/1.9V anymore. I may move it back to the blue slots and see if that helps.
-bZj
down8, by chance I found something that might be relevant, see this memory from OCZ:
http://www.ocztechnology.com/product...0_gold_edition
Notice the note they put there regarding the speed:
1600MHz DDR3*
* Test specifications are based on dual core CPUs only. If quad core CPUs are used, higher bus thresholds and core arbitration may necessitate lowering the memory frequency below tested specs.
This could explain something I noticed while trying settings here the past days, when I tried 333/1333 the computer didn't take as much to boot (from none to max of 4) with the default 400/1600 can be much worse.
Since I want to keep 1:1 ratio I am still trying to see what if any settings for 400/1600..... but if you notice most people running the higher memory frequencies here are using dual cores.
I hope that when they release the official memory (JEDEC) PC3-12800 they will accommodate the quad cores... otherwise we would be in the same joke.
well, been a while since i had the time to visit...looks as though things with the newbies are falling apart and gradually being put back together again as usual (dont want anyone to feel like they got cheated because they didnt get to "experience" this board!).
anyway, updated BIOS, and still cannot lock multi.
sean
and GUEST, i think that we are the only 2 with q6700's in here, so when i get some time i will look at what your settings are and see if i can help. right now i am at 362mhz, or somewhere around there, and will probably stay there for now...i have a testy core and really have to crank up the juice to keep it stable.
I have a Q6700.
Nerea, are you using:
http://www.memtest.org/
http://www.memtest86.com/
http://hcidesign.com/memtest/ ?
That is interesting about the speed being reduced while running a Quad. I have an E2180 I may be able to test with later, but I don't want ot tear down both machines right now (again).
-bZj
Hello (I think I solved the problem here) :)
I am using http://www.memtest.org/
Well today I tried to set just in case 400/800...... the computer enters an endless boot loop with that, I needed to use the jumper. There I set the 333/1333 which used to more or less work (would boot after a very few tries)
Not anymore, it took it way too long and by then I was thinking in getting any other memory since this one was not compatible with this X48. But once I managed to enter the BIOS I tried leaving the reference frequency in auto, still keeping the memory frequency at 1333. It worked at the first try..... but the computer was left with memory at 1333, tried just in case with memory frequency at 1600 (remember leaving the reference frequency for it in auto)
To my amazement it works, works at first boot, cold boot, after turning off PSU etc. I tried booting several times and every time it booted right away.
So no idea if that might work there, but try leaving the memory reference frequency in auto instead of 400 or 333.... while setting the memory frequency where you want it. See if that works.
Default/1600 locks - jumper city.
400/1600 => 1333.
Weak.
(Sorry about the lock :S)
For 400/1600 and getting 1600.... you need to have the FSB oc to 1600 mhz too. But I have no clue about FSB oc nor CPU, when I build a computer I keep it for a long time..... from 4 to 5 years. So I go for something bleeding edge when the time comes, but I try not to force stuff because I want it to last while being very reliable (my current computer works perfectly after all that time, only upgrading the graphics card) and so did the one before this one. :)Quote:
400/1600 => 1333
I always get Intel mobos because of the reliability, and for me was very strange to deal with this memory problems. During these days I tried to see the very basic of what is what in the memory settings, but since I don't oc the system I am the last person to advice on that.
But check this out if you still didn't, is the memory guide for the BT posted by gtj. There you see what numbers you need to get a target memory frequency working.
http://www.peakin.com/x38/calculator.html
I know tho, that computers default to the processor FSB and that is the barrier you need to see at to what point can be overcome. With your processor defaulting to 1066, the mobo at 1300 might be difficult to get the system stable at 1600 mhz, Intel mobos aren't too good for oc.. yet might be possible, the other folks here know how to oc.
Here are my Everest results for my X48
Hi everyone, I just got this board, and i just found this forum. I'm not really intrested in OCing it that much, I guess I might be having the same problem with the FSB speed and the memory, i set it to 1333 and 9-9-9-24, 1.7v. and it works now, however when I did that, and pretty much anytime I change any settings with the memory, (havent messed with the cpu until i get a better HS and Power supply), I cant get anything to display, and it takes what seems to be a random combination of changing the jumper, and for some reason unplugging and plugging back in my video card. It just seems to be a very finicky board as far as settings go, but I know its a solid board and it had everything I wanted in a board.
Heres my build
I flashed the bios to 1610 or whatever the newest one is.
CORSAIR 2GB (2 x 1GB) 240-Pin DDR3 SDRAM DDR3 1333 (PC3 10600
Core 2 Quad Q9300
Geforce 6800 GS (Im replacing that soon!)
Basically if anyone has a similar build, I wanna know whats a fast, stable setting for my memory, and what the best way to check that it is indeed running at 1333, I dont really want to make it unstable cause Ill probably be running it most of the time.
Pontius, I wasn't oc either and as you only hoped to get the memory working at rated speed with the rest of the system without problems. Nothing worked for me until yesterday, by chance solved I it by setting the reference frequency in auto, while setting the memory frequency at the rated speed. In your case try auto in the reference frequency instead of 333, and set 1333 in the memory frequency. Also check the voltage.
Besides this, if your system boots without display can be a problem with the graphics card or PSU as you are guessing.
At present I have E6320 1.84 GHz CPU 1066 overlocked with DX38BT:
1. Host Clock Freq 498 MHz
2. Multiplier 6x
3. Memory 400 MHz
The BIOS and Intel Desktop software reports the CPU running at 2.99 GHz.
Windows reports 2.32 GHz
BIOS also reports a different voltage which is setup correct.
Intel Desktop reports:
RED CPU Volt .80
RED MCH Volt .80
RED Power Supply -
Help, the system seems stable, but something is wrong!
yeah now that you mention it, its more than likely my power supply. everytime ive seen a computer not display something 9 times out of ten its the power supply :P oh well the new one comes in on monday. I'm pretty shure my memory is running at the right speeds now, I know I'm using less memory in windows now so it must be :P thanks
great thread btw I'm super impressed and a little intimidated ha!