Printable View
Quite possibly the NF4-D is using different revision numbers as opposed to the NF4 Ultra D, causing the NF4-D to have the same changes as the lastest AD0 NF4 Ultra D's (in terms of physical componantry) but with a revision number that is 2 revisions prior to the latest seen on the Ultra. This could have been due to the NF4-D being launched slightly later than the Ultra D or requiring less bios changes over its development as opposed to the Ultra D.
I guess the reading from the probe is interpreted in a different way by this board or sth. Really ard to tell, but there is more to it than the probe itself, which is a quite passive component. The logics used to read the probe and eventually the hardware used for it plays a role :)Quote:
Originally Posted by g0dM@n
i love my AD0. my chip used to need 1.51v for 280x9, when i was testing it. the vcore would flux up and down. now with my new AD0 Ultra-D from the egg, vcore is PERFECTLY stable at 1.408v, and 280x9 id chugging along like a champ.
Reply from DFI Distro in my country:
"Hi ABCDE, thanks for your interest in our products.
As far as I know, AD0 is a bios revision, not a mainboard revision - you should be able to flash any board's bios to the most current, thereby alleviating the need to purchase with that bios already installed.
Hope this helps;
Thanks, have a great day!"
Placebo effect on you guys lol !!!
I agree with LV.
Oh. Why are the components differently placed on our motherboards then?Quote:
Originally Posted by GMX
As i said, AD0 might be a bios revisison, but there IS a new board revision too.
It's funny how when you ask them they don't even know. We know more than the people we ask questions to.Quote:
Originally Posted by SlackeR
I need more HTT and I'm getting ready to make the jump from my AB0 Ultra D to a AD0 just on the chance that I can get some higher speeds out of it.Quote:
Originally Posted by g0dM@n
Desperation setting in. Am I wasting my time here? Newegg will be shipping today.
I wanted to buy the Expert board just b/c of higher speeds and stability, but I couldn't justify $200 for it. I also don't need SLI, as I'm using an ATI card.
If the AD0 really performs almost as good as Expert, then I'm going for it too... just need more proof. :D
Also, why go Expert when you can still mod these to SLI? I have had an extra SLI bridge anyways, so...
i've got a AD0 too :D , the highest htt i could achive was ~295 :( :stick:
You're not giving me a high level of comfort here with that statement. Anybody else have any answers?Quote:
Originally Posted by Turrican
I've read your other thread (about hitting 300+ htt) and i've got the same problem with my board. at 300, the board won't boot. :(Quote:
Originally Posted by Rapier
Maybe we can work together to solve the prob. :)
c u
I run my opty @ 300x10 4x HTT without problem :fact:
You guys are hitting a wall at 300? That's not like the DFI NF4...
Are you guys running low CPU multis? Are you running the x3 or x2 LDT?
Also, you should try upping the LDT or chipset voltage. I've had gains with upping the LDT voltage before.
Another biggie is a RAM divider...
i.e. My Opteron 148 would not boot beyond ~285fsb with the x10 multi on the 166 divider... when I ran 1:1, it did 300x10 with EASE...
You guys may be overlooking things... Can you verify? :D
1. I'm running it at the max 10x multiplier.Quote:
Originally Posted by g0dM@n
2. I've played with all the voltage options core, LDT and chipset. Nothing seems to make a difference.
3. I'm running Geil PC4400 ram on a 166 divisor. The stuff is rated for 275 fsb so that shouldn't be an issue.
Puzzled (with AD0 on the way, now)
Better jumperless ram voltage options, jumperless SLI configuration (a real benefit if you like bench comparisons between SLI and non SLI), and, most importantly, a MUCH better spacing between the PCI-E slots that allow better SLI cooling solutions to be used, and a 3rd PCI slot that is actually usable when the board is configured for SLI using double slot width cards.Quote:
Originally Posted by g0dM@n
These improvements are worth every dime for the hardcore users running SLI and water cooling.
Refer to post #177.Quote:
Originally Posted by SlackeR
I believe if more people referred to whether they are using a Ultra D or NF4-D when quoting the revisions, this may help resolve the discrepencies we are seeing. LV & Wai Wai's identical AB0's have both been NF4-D's not Ultra-D's.
Not too sure if that first Ultra-D AB0 LV quoted though has the AD0 layout we are looking for, maybe this can be verified.
Quote:
Originally Posted by g0dM@n
i'm running LDT x3. LDT is set to 1.40V and Chipset Voltage is set to 1.70V.
I'm using ram divider 166 with my ocz vx 4000.
I use cpu-multi 10, but it also doesn't work with 9. :(
with a htt over 295 windows won't boot. it always freezed during loading.
max htt is 290 now.
c u
This with the Vapo? Sure your CPU doesnt get effected by the cold bug?Quote:
Originally Posted by Turrican
yes it's with the vapo :) . i don't think the cpu get effected by the cold-bug, because it's able to run 264x12=3168 24/7 stable. Why should there a cold bug with 300 htt? :confused:Quote:
Originally Posted by MrRevhead
c u
Wow, I overlooked all of those. Good points. Well, I guess my point kinda does make sense when you think about the whole SLI thing. If you're not going SLI at all, then it's kinda worth it... only thing is that the ram voltage options are awesome on the Expert. :DQuote:
Originally Posted by mdzcpa
Dude, I SERIOUSLY think that it's your divider that is causing you your problems. Before my opty170, my opty148 was my main rig and I had a 144 as my 2ndary. They 148 would run my 180 divider with my bh5 ram without a problem. I was running 275-280fsb x 11, but my ram was only at high 230s that way (~240mhz @ 1.5-2-6-2). That was NO WHERE near what my ram could do, so I went ahead and used someone's divider/fsb program (I forget who made it, but there is a prog that I was using that helps you out where you plug in the cpu freq you want and ram freq and it gives you lists).Quote:
Originally Posted by Turrican
I also read up about how the dividers are calculated exactly and figured out that at 300htt and 166 divider that the ram does 250mhz, which was almost perfect in my case.
So... I went ahead and tried the x10 multi. The board would NOT even get to windows logo without rebooting at anything above ~285HTT with the 166 divider. I took a guess and decided it was the divider that was screwing it up. Unfortunately, I had sold all of my TCCD before purchasing the 148 (I had the 144 for a little bit before the 148 and had no issues with UTT or BH5).
I figured that I'd try to use no dividers with the x10 multi to prove a point. I went ahead and bought some OCZ Plat Rev2 3200 off ebay. I threw those suckers in at 300x10 with a 1:1 multi at 3-4-4-8 and wabaam!!! I made it to windows without a hitch!
I ended up stabilizing 300x10 2.5-4-3-7, and probably coulda pushed more, but I definitely proved to myself that the 166 divider in combination with the x10 multi wasn't working.
Conclusion: always test your dividers and play with them, especially when it comes to figuring out the max HTT of your mobo. Another good thing to test is even LOWER multis (maybe x7 or x8) cuz they may work differently (better) with dividers.
You cannot claim that a board doesn't do a certain HTT without trying out all of these factors... don't forget, LDT voltage and chipset voltage may help as well.
I know you said that you tried different voltages, but maybe if you can't borrow a TCCD/5 set from someone, then you can try lower CPU multis with the dividers and see if that helps. Not for all, but dividers are REALLY picky with some setups...
True. I have AD0 Ultra-D for the record. "AD0 Layout".Quote:
Originally Posted by MrRevhead
1st: thanks for the tips mate :toast:Quote:
Dude, I SERIOUSLY think that it's your divider that is causing you your problems. Before my opty170, my opty148 was my main rig and I had a 144 as my 2ndary. They 148 would run my 180 divider with my bh5 ram without a problem. I was running 275-280fsb x 11, but my ram was only at high 230s that way (~240mhz @ 1.5-2-6-2). That was NO WHERE near what my ram could do, so I went ahead and used someone's divider/fsb program (I forget who made it, but there is a prog that I was using that helps you out where you plug in the cpu freq you want and ram freq and it gives you lists).
I also read up about how the dividers are calculated exactly and figured out that at 300htt and 166 divider that the ram does 250mhz, which was almost perfect in my case.
So... I went ahead and tried the x10 multi. The board would NOT even get to windows logo without rebooting at anything above ~285HTT with the 166 divider. I took a guess and decided it was the divider that was screwing it up. Unfortunately, I had sold all of my TCCD before purchasing the 148 (I had the 144 for a little bit before the 148 and had no issues with UTT or BH5).
I figured that I'd try to use no dividers with the x10 multi to prove a point. I went ahead and bought some OCZ Plat Rev2 3200 off ebay. I threw those suckers in at 300x10 with a 1:1 multi at 3-4-4-8 and wabaam!!! I made it to windows without a hitch!
I ended up stabilizing 300x10 2.5-4-3-7, and probably coulda pushed more, but I definitely proved to myself that the 166 divider in combination with the x10 multi wasn't working.
Conclusion: always test your dividers and play with them, especially when it comes to figuring out the max HTT of your mobo. Another good thing to test is even LOWER multis (maybe x7 or x8) cuz they may work differently (better) with dividers.
You cannot claim that a board doesn't do a certain HTT without trying out all of these factors... don't forget, LDT voltage and chipset voltage may help as well.
I know you said that you tried different voltages, but maybe if you can't borrow a TCCD/5 set from someone, then you can try lower CPU multis with the dividers and see if that helps. Not for all, but dividers are REALLY picky with some setups...
2nd: i've tried allmost everything to get the board stable past 290 without success. i've tried it with multi 10,9,8,7 and 6 and it's always about the same :(
i've also tried the dividers 100/133 and 166 with different ram settings + different LDT & Chipset Voltage. LDT Multi was always set to 3. :slapass:
Could it be that i've got a bad mem-controller on the cpu, so that ~290 would be the max.? :confused:
I think the best i can do is, to sell the vx and get some tccd. Could you recommend some good sticks ? (G-Skill, or whatever else? )
thanks in advance Turrican
If you didn't live overseas, I woulda lent you one of my TCCD sets. I cannot really say that you should replace your current set with TCCD b/c it's likely that it wouldn't solve your problem.Quote:
Originally Posted by Turrican
What you're going through is the reason I bought TCCD again. I used to have gigs and gigs of UTT (after I sold all of my TCCD), but after I came into the problem that you were having, I ended up convincing myself that I needed a set of TCCD around as well.
Maybe you could find a 256mb stick of TCCD for <$50 and just have that around... see if you could do that... :D
Unless, of course, you can find someone to LEND you a kit.
Hmm i did have the same problem but i found a cure for it :), before i had a winchester(orginal multiplier was 9x), i could not run over 260fsb with multiplier 9x, used 8x to i got my 146. The 146 got orginal multiplier 10x i got same problem with it 10x didnt work over 260fsb but 9x did work on the 146.So i tryied to find out the problem"settings in bios and more" . Then i reflashed the bios and after that all multipliers and dividers worked fine :)Quote:
Originally Posted by g0dM@n
BIOS fixed a divider problem?Quote:
Originally Posted by Lastviking
Well, my problem was either on the 702-3, or the 623-3... don't recall.
Turrican, try a diff bios like the man says... :D