• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

1T + divider = no windows

Joined
Jan 15, 2004
Messages
705
ok my system doesnt seem to want to go into windows at 1T. im forced to use 2T unless i want absolutely almost no overclocking.

with 1T, im unable to go into windows with a memory divider. my ram the gayest PQI turbo 3200 DAH, will not overclock whatsoever.

only with 2T am i able to use a memory divider. will the 2T give me a performance hit?

i cannot use a memory divider because im unable to enter windows if i do. without a divider, i cant go above about 222HTT and even then my memory is REALLY unstable.

specs are:
dfi lanparty ut nf4 ultra-d
3000+ venice
1gb (2x512) pc3200 PQI turbo; the DAH sticks
ram is rated at 2.5-3-3-7 by mfg, but cpu-z in the spd says its 2.5-4-4-8. i had them set at 3-4-4-8 at the time of OC...
enermax psu; eps12v, 24pin, about 33amps on each rail.

the venice i just recieved today. i had previously been using a winchester which had a max OC of 260 * 9 @ 1.64v with memory at 150 divider and at 1T settings.

ive ram memtest on my ram at stock speeds and at OC speeds. it was error free up to about 230mhz but as i increase speed starting from 200mhz, i experience problems in games, 3dmark05, and pretty much everything that requires loading... so im forced to run at 200mhz or less...

to be more specific, this is what happens...

i boot up my comp at 222 * 9, 3x htt multiplier, it starts to load windows (see that screen with the logo) once it gets to the desktop (not everything is finished loading yet), i see the desktop for about half a second before it reboots. ive tried upping my v core to about 1.65v and still problem persists.

not this did not happen winnie but i was only able to get 2.34ghz out of it.


anyone know whats wrong? i bought the venice expecting a nice overclock and now iv ran into this fvcking stupid 2T thing.
 
I have the same processor and the same board - my first thought is to ensure you've update your bios to the most recent. Either the DFI official (3/10 I believe) or if you can find it (Sorry, I can't recall where I found mine so am little help to find the link) one of the "beta" or 3rd party bios's. Reasoning behind it is both better ram/spd interpretation which may fix that ram issue - plus better / enhanced venice core support.

Perhaps someone else can clarify the bios situation... but at the very least try the 3/10 bios and see if it helps. I do recall (wherever the heck I found those 4/xx beta bios's) that the beta bios text included ram compatibility though. They were one a 3rd party site (found through google) - not DFI's official site. DFI forums also don't seem to have any info on it with a quick check.

I had the 2T issues with my old pc2700 ram - but once I got pc3200 corsair value-ram it was fine. Still hard to narrow down timing, but works fine at 1T.
 
I'm having very similar problem with DFI ut sli dr and twinmos memory haven't tried 2t yet though.
 
Autochthon said:
I'm having very similar problem with DFI ut sli dr and twinmos memory haven't tried 2t yet though.

Again - what bios revision? I'm going to go google diving and see if I can find those beta bios's again.
 
Googled for "dfi lanparty ultra-d beta bios".

First result was this. That last bios on the list is exactly the one I'm using. I've heard mention of this Oskar Wu character on the DFI official forums, but have no idea who he is.

Either way, the bios's work. YMMV for fixing the 1T/2T timing issues with ram.

edit - added the link in my sig - also read the forum post, with the 05/10 bios's there's 3 revision with no specificiation as to what does what - but there's speculation that there's TCCD compatibility differences with focus on some winbond compatibility issues on one of the three. I wish I could find the source for these with some documentation instead of these crummy 3rd party sites.
 
Thanks for the link. I tried the beta bios it didn't seem to change anything obvious in my case but didn't appear to hurt either.
 
Lukano said:
I've heard mention of this Oskar Wu character on the DFI official forums, but have no idea who he is.
Oskar is the lead bios engineer at DFI. remember back when ABIT was awesome? guess who was the lead bios engineer there at the time? :p

for the beta bioses:
xxx-1 = optimized for tccd
xxx-2 = for bh-5/ch-5/utt
xxx-3 = the generic bios that will work well with anything
 
alright, ill give the 5/10's a try. i was using 3/10 before when the problem was there. then i switched to 3/16, still no luck. ill report back in a little bit...

only thing that bothers me was i didnt have this problem on the 3000+ winchester i was using.


edit: ok i tried to flash it and it didnt work. I copied the unzipped files onto the floppy disk and tried to boot off of it and it kept saying something like "please remove media and press any key to continue"

someone help me out?
 
peterhoang2002 said:
alright, ill give the 5/10's a try. i was using 3/10 before when the problem was there. then i switched to 3/16, still no luck. ill report back in a little bit...

only thing that bothers me was i didnt have this problem on the 3000+ winchester i was using.

edit: ok i tried to flash it and it didnt work. I copied the unzipped files onto the floppy disk and tried to boot off of it and it kept saying something like "please remove media and press any key to continue"

someone help me out?

it's a horrible horrible suggest (I'll probably catch a fair amount of flak for admitting it) But I use winflash and the raw bios to flash mine. Flashed it 3 times now without a problem - just make sure you're not overclocked (so no risk of it crashing mid flash) if flashing from windows.

the fact that it didn't happen with the winny leads me to beleive it's bios support for the venice and/or if you got new ram that too.

Now for my own personal side question - TCCD mean?
 
i have a few 512 sticks (other than the pqi) lying around. i will try to use those and maybe try just one pqi stick at a time to see what happens.

ill also try the winflash thing. ill tell you results after im done.


update: ok, flashed bios to 5/10 using winflash. rebooted at default settings and got into windows. good so far right? so then i went back to bios, return to last stable overclock settings (2t). boot, got into windows, ok good. when back to bios and changed memory to 2t, booted got into windows for 1 second and crashed. not so good and same as before. ok after that i went back to 2t settings and booted only to find that the last crash had corrupted one of my system files and now windows failed to boot. so right now, im reinstalling windows to try again with just 1 stick.
 
awesome!!! this new bios fixed my problem!!!

after installing windows, i started fooling around with the memory slots try different combinations and i found one that the computer likes. placing them in the yellow slots (i think 1 and 3) allows me too boot into windows at 1T. hmm maybe this bios will also now allow me to overclock my memory without the problem i described in the last post of this thread...

update:
ok well looks like the ram is bad. it boot into windows but after 5 mins, it crashed once again. i tried to boot once again to run 3dmark05 and once again crashed during the loading screen.

lowered my ram divider making the ram run at 183mhz and kept 1T settings, then ran 3dmark05 and looks like its stable here...

i think im gonna RMA these 2 sticks as they dont seem to be running well at what they are rated at...

update:
ok i just tried calling the pqi company and turns out the number given on their website is no longer in service or at least the message i get when calling them says so :mad: :mad:
 
i tried the 310P "performance" bios, and i have to tell you, its gotten my winchestor up from 2.3 to 2.5GHz so far, which doesnt sound like much, but going from a semi-stable 2.3 to a solid 2.5 is pretty sweet. i would reccomend it for anyone who has a similar setup as i do, as it has opened quite a large overclock to me!

thank you Lukano very much for the link!
 
Why are you keeping 1T? Go 2T and raise your fsb. You are not going to loose much (anything noticeable) at all.
 
Why are you keeping 1T? Go 2T and raise your fsb. You are not going to loose much (anything noticeable) at all.

yea, itll only cost you about 20% of your bandwidth.
 
actually, 2T doesn't effect bandwidth much from my testing, just latency by about 10%
 
I know this may not be the best place to post this - but I read mention of someone RMA'ing their ram as it wouldnt' do the timing advertised.

I've got Corsair Valueram (bought from ncix.ca) that had no rated speeds in the description nor on the package. I'm coming to the conclusion that this is relatively poor ram, but as it has no "Advertised Speed" I can't think of or find any justification to request an RMA. I really am not happy with it - and would like to bump up if possible.

So does anyone have any experience RMA'ing ram with NCIX.ca and/or with Corsair direct for no really valid reason other than being unhappy with it?
 
ok new update on this. the ram may be the culprit but it is DEFINITELY not the ram i thought i bought. allow me to explain...

i went to frys today to pick up a wireless network extender and coincidently i saw someone holding a pack of pqi ram so i went up to him thinking that i may warn him about it to do him a favor but then i noticed something. the ram was the SAME EXACT model as my ram, or at least what the numbers and model number said. what wasnt the same was the heatspreader. suspecting that they guy i bough the ram from switched some ram from me, i went ahead and bought a pair of the ram i was suppose to have from frys to compare...

here are some pictures that will do better explaining...

new ram in new sealed box. ahh fresh ram, looks nice!
real_nib.jpg


new ram up close with the reciept:
real_receipt.jpg


now, heres where it gets interesting.
comparison.jpg

new ram is on the left, fake is on the right. noticed the difference in the color of the heatspreaders? now also noticed the things i boxed. on the new ram, the pins are all going the same way when the label is facing you. on the fake ram, the pins are different... clearly the person i bought it from tampered with this. he had taken off the heatspreaders (or the label) and carelessly place the heatspreaders (or label) on the wrong side...

again, another close up comparison so u can see the difference in color...
comparison2.jpg


this time, on top of each other
comparison3.jpg



heres pqi's website: http://www.pqimemory.com/spec.asp?link=/dc/PQI3200-1024DAH.htm
notice, it says "Gunmetal" and not mirror black.

so ive come to the conclusion that this has been tampered with. i bought this off an anandtech forum member so i will try to get a refund. if he refuses, i will spam this ass's username on every computer forum i know, but for now i feel he deserves a chance to make it right.

But of course, the question here is 'does the real ram make a difference?' well i just got the dreaded BSOD that no one likes so im gonna reformat once again and ill update on how it goes with the new ram.
 
I hope you realize that RAM resold by companies such as Corsair, PQI etc... are not always the same chips and board even for the same model and series. I highly doubt someone would go out of their way to cheat you out of the RAM. It's rated at PC3200 2.5-3-3-7. If it doesn't go past that, but it meets those specs, then you really have no reason to complain. Overclocking is mostly luck of the draw, you're gonna get some pretty average hardware that doesnt match up to what everone else is getting every once in a while. Just suck it up and move on.
 
peterhoang2002 said:
ok well looks like the ram is bad. it boot into windows but after 5 mins, it crashed once again. i tried to boot once again to run 3dmark05 and once again crashed during the loading screen.

lowered my ram divider making the ram run at 183mhz and kept 1T settings, then ran 3dmark05 and looks like its stable here...

i think im gonna RMA these 2 sticks as they dont seem to be running well at what they are rated at...

doesnt run at pc3200, so i guess i have a reason to be mad?
 
peterhoang2002 said:
doesnt run at pc3200, so i guess i have a reason to be mad?

If it does not run at PC3200 at STOCK speeds for motherboard, and CPU - then yes. Otherwise, back to the "luck of the draw" for overclocking stability/performance.
 
peterhoang2002 said:
doesnt run at pc3200, so i guess i have a reason to be mad?

You were running it at 1T, from what I understand. It's not marked for 1T operation.
 
Back
Top