DFI LanParty 875P-T LGA775 motherboard

[Mauri Antero]

I also know that the board got more potential that it has shown me.

First thought that it is nb issue but doupt that after changing the chewing gum away from nb heatsing and put new thermalpaste in. PM:d Fugger what bios he used when he drove that 6GHz , so far tried the last 3 bioses that was in DFI site.

[Angry_Games]

also realize he was using phase-change or some other sub-zero cooling that normal users couldn't possibly compete with, and was not using 1:1 either lol.

[Mauri Antero]

also realized that but still that over 400fsb wonders me. Dryice is my next cooling, as soon as possible. That goes also to -79ºC and cools nb too.

[Angry_Games]

heh im still on air cooling with a half-complete Corsair Hydrocool EX rig behind me (and i've never touched watercooling nor subzero stuff)

[Mauri Antero]

Well now running with air Zalman 7700 Cu on the pros cooler, easilly goes to 4300Mhz. Now just trying to find optimal settings so that I can waste some CO2 and later if CO2 drives goes nice ln2.

[MrBean]

Hello Guys,

Been a while since I replied here....anyway, I have my 875P-T setup up and running, and are relatively happy with the board.....however, that said, I am experiencing a few Bios funnies with this mobo, some stability issues, and some overclocking issues.

 

I am no newcomer to the overclocking world, having been involved with watercooling and overclocking since 1997, and the last 2 years with phase-cooling.

 

I managed a healthy 15579 in 3dmark03 with 6800U, which places me no1 in the world for this class with WHQL certified drivers, and 3.5.0 fix on the Benchie. Which aint bad.

 

linky:http://service.futuremark.com/compare?2k3=3441900

 

3dMark01 I am no7 for 6800U class, at 30940, but the cpu is holding me back. One moment it will overclock to 4.2 gig on this board, the next it will fail benchies at 3.90.

 

Linky: http://service.futuremark.com/compare?2k1=8291082

 

So, to be completely honest, it is really difficult to get the feel of this board, and I have to say I am a bit disappointed at DFi not solving the Bios memory issues.

 

To be fair though, I need to remove the NB cooler and apply a better thermal paste to it...but that will not really solve any mem stability issues

 

I will keep you guys posted of developements.

 

Br,

MrBean.

[BigStan]

from MrBean ~

I have to say I am a bit disappointed at DFi not solving the Bios memory issues.

from Angry ~

1. that the 5:4 divider will indeed work like we expect it to. We have found only few modules that correctly function on a 5:4 divider. Every other Intel P4 board works flawlessly with 5:4 and this one should be no different.

This BIOS issues runs deeper than one might think. It's not just that the 5:4 divider doesn't work. It will corrupt a Win install. On my IC7Max3, my Adata mem will run at 278fsb 1:1. When I run it on this 875P-T, if I use the 14 multi to to to 275 1:1 it runs fine. When I go to 280 1:1, it gets a little flakey. I would expect it to. But when I change the divider to 5:4, instead of booting up, it corrupts my Win install. The first time that happened I thought it was a fluke. The second time, I sat up and paid attention. The third time, I deliberatly set it up at 280 5:4, and sure enough it cr*pped out my Win install again. This is a consistant repeatable event. I mean corrupt to the point is needs a Win repair install to get going again.

 

I also think that the AGP/PCI lock starts to fail when you get above 275fsb. I can boot at 280 1:1 and run a few benchmarks. But I cannot boot into 285fsb 1:1 without changing the AGP/PCI frequency to 73/36. It is unstable, but it will boot into Windows.

 

Angry ~ that fact that I can get run at 275fsb dissproves you and Bigtoe about the fact that this board is limited to 250~260 fsb. It will do more than that. But these BIOS issues at the higher fsb ranges are going to lead to a slow death for this mobo. There will be a lot of other OCers like myself and MrBean that are going to stumble on them. And that will eventually lead to a bad reputation for this mobo if they are not fixed.

[MrBean]

Hello Everyone,

interesting topic, this one .

 

This board has so much going for it, a real overclockers dream in terms of hardware design, good vregs that stay ice-cold under overclocking, nice layout, I can even get my large paws in there without a issue.

 

So, basically it is just this bios nigglies, and if they can be sorted, this will be thee board to have....what concerns me, as I have mentioned, is not really the memory...or maybe indirectly it might be.....

 

One moment I can overclock my 3.4EE to run 3dMark03 at 4.16g on the cpu, 1.775Vcore, then the next it fails at 1.825Vcore and 3.90-3.97g. There is a problem with the multipliers as well, although to be fair, it may be cpu related.....this is an engineering sample with unlocked multies, and I have mixed results at different mutlipliers.

 

What is weird is that the 16X multi on the board is not that stable for me, while 14, 15 and 17 is doing a good job.

 

But, Angry Games said he will chat to the Big Guns in Taiwan.....let's hope they respond sooner rather than later with a very nice Bios update that will sort some of these issues.

 

Have a nice day, talk to you all later.

 

Br,

MrBean.

[Mauri Antero]

Well think the issue is with that 5:4 ratio drive, will try today with -78ºC dryice. Max with Abit AS8 is now 5144 and SuperPi 8M and wiching to go higher with this board. Will post the results after the tests.

[Angry_Games]

Angry ~ that fact that I can get run at 275fsb dissproves you and Bigtoe about the fact that this board is limited to 250~260 fsb. It will do more than that.

 

couple things to remember

 

1. your 275 is atypical. Ive been playing some more and can hit 280 1:1 boot, as can Bigtoe with loose timings and the 'right' setup.

 

most on average, will not hit this mark. A few of us will have the proper equipment and knowledge and experience to break 280 consistently.

 

this doesn't mean we dont care about those hardcores that are after the +260 ranges. It just means that on a scale, there are very few of us that will be trying for these high ranges.

 

It doesnt make us any less important, but as everyone knows, no overclock is guaranteed and everyone's results may differ. We have already witness a user hitting 6Ghz, so we know it can be done.

 

Its more of a question of getting used to the board, knowing what it can and cant do (without prior experience from other boards coloring your expectations).

 

I myself am feeling like the more that i play with it, and try out different settings and modules, am getting a much more comfortable feel for it and how it reacts.

 

This is no different than when I first played with teh NF2. Our DFI NF2's are not like other NF2's...they have their own personality and it has taken many more and better skilled overclockers than me to come to the realization that it was all about learning the particular board and figuring out where it could be manipulated and cajoled and sometimes even threatened (lol) than trying to apply all prior skill and knowledge to the overclock of the NF2..

 

what that means is if you go through and ask the majority of our NF2 owners, they will tell you a very eerily similar story...about how they were doing this or that high overclock on their and expected to just ramp right up with the LP NF2 or Infinity NF2.

 

didnt quite work out like they expected. Some went crazy and jumped out a 30th story window as the madness finally got the better of them. Some came to forums and threatened to kill me, my family, and all past, present and future genetically similar relatives. Some even threatened lawsuits against DFI for making defective boards hehe.

 

after the smoke cleared and the fires died down to a good solid clean-burning blaze, we all took a collective sigh of relief as we compared notes and stories and found out that these mean NF2 DFI boards were extremely clockable, and extremely friendly to tweaking and diddling with.

 

but the price for that was learning the nuances of the particular board, and more importantly the bios and the hardware you were putting to the board.

 

 

But these BIOS issues at the higher fsb ranges are going to lead to a slow death for this mobo.

 

doubtful. You have your opinion and I dont mock it or call you out on it at all. But before you start the parade of death you should realize that bios development is an ongoing process, and one customer not satisfied that his board will only overclock to 270 instead of some other number is definitely not indicative of whether the lifespan of the board will be long or short. Neither is 2 or even 10 customers.

 

its really a moot point as I have already explained to TW that the bios needs some work on it still, and they do agree. Neither they nor I can guarantee you or anyone a time frame as to when a bios will be released to address this issue.

 

And keep in mind that the 5:4 issue is a much more critical issue to be addressed than any overclocking issue. Sure 5:4 aids in overclocking but its a much more critical event to have a 5:4 failure than it is to only hit 270 instead of 280 (and just because you did it on another board doesn't guarantee anything on a different board).

 

I've stated to TW that the 5:4 issue is something that must be addressed before anything else because that is the truth.

 

 

There will be a lot of other OCers like myself and MrBean that are going to stumble on them.

 

you guys and at least 20+ others that have this board and have used their knowledge and skill and experience and encountered some issue or another. This is why we are here at this forum to hear what you guys have to say. Normal companies would tell you to shaft off because overclocking isnt supported.

 

we say overclocking is not supported nor guaranteed, but we want to hear what you have to say so we can try and alleviate any overclocking pains you might have. In the end we don't guarantee that you will get any overclock, nor do we guarantee that we can or will fix a specific issue that deals with overclocking (that doesn't have anything to do with standard functionality of the board @ intended, default speeds).

 

sounds trite maybe, but its just the truth. I care, RGone cares, Jacky cares about overclocking and making sure our boards are great clockers. But in the end, we simply cannot guarantee that we can get any overclock out of any product. We can only guarantee that it will perform @ it's intended, default speeds/settings.

 

And that will eventually lead to a bad reputation for this mobo if they are not fixed.

 

i've heard it before about other boards, from DFI to ASUS to Abit etc. As I stated, i care about overclocking and making our boards the best but I'm not really concerned that this boogyman of a 'bad reputation' will come scaring us into oblivion.

 

Truth is that whether you or me or anyone complained about any features or issues on overclocking, the boards would sell at a given pace.

 

Before you go thinking that is downright arrogant, think again. Abit and Asus and Giga and MSI and Shuttle and every other mfg knows the truth of this. They know that their boards are exceptional at default intended speeds.

 

they know that their boards perform very well at overclocks.

 

they know that there are a few that are going to go for broke and may or may not get broken lol.

 

in the end they all know that the majority of users are going to be happy with a decent overclock (out of those that actually care about overclocking).

 

Abit only got you 254FSB 1:1? Thats actually a good clock.

Asus only got you a 248FSB 1:1? Not bad not bad.

Giga's board gave you 266FSB 1:1 with much tweaking? Still within what overclockers (and head monkies of mobo companies) consider very good clocks.

 

what I am trying to tell you is this:

 

you (and me, and a few others) are complaining that 260FSB is not good enough. Unfortunately, THAT is not good enough. It doesnt matter if it is Asus, DFI, Abit etc. We are a very small micro percentage of overall users and we are complaing about something that is not even covered or warrantied or guaranteed...only expected, and that expectation comes from the customer (who, again, is a very micro percentage of overall sales), and that customer might or might not have the skill to achieve such overclocks as he thinks the board is capable of.

 

the complaint that 5:4 hoses windows is nothing extraordinary, as that is to be expected from a failure to correctly function at a ratio.

 

But it is a very relevant concern and is something we will address first before anything else as those ratios are supposed to function correctly...and looking at other boards within the same chipset line, if this one does not function correctly when all other 875 chipset boards (regardless of mfg) do...then we have a problem that must be fixed...and DFI-TW is working on it as we speak.

 

but don't get your hopes up high that you will get this magical bios that will allow you the 300FSB 1:1's that you expect the board to give.

 

Heck, I expected to hit 300FSB since my 875B hit 291 with the same exact hardware minus cpu (since its an LGA775).

 

But i can't see myself rolling out the deathwagon for the board if all I ever get stable out of it is 275, or 265, or 291 or anything.

 

If the board wouldnt even do 220FSB 1:1 then I'd be worried and my tune possibly might be different. But we are not talking about 220, we are talking 250, 260, 280, etc. Something that might be expected from some users, but overall is still quite a luxury from any board (and one that very few take advantage of as majority of users are still more interested in stable stock speeds than catching their hair on fire from a crazy overclocked cpu static discharge lol).

 

This BIOS issues runs deeper than one might think. It's not just that the 5:4 divider doesn't work. It will corrupt a Win install.

 

as i stated, this isnt any shocking new news. This is to be expected when the ratio/dividers arent working correctly. I'm disappointed that they dont work, and am pushing hard for them to hurry up and fix it. But it isnt a big shock to have my winstall corrupted. Its just disappointing that its not working properly at this time. Disappointed enough to annoy the heck out of anyone in DFI that will listen so we can have a fix for it.

[Angry_Games]

One moment I can overclock my 3.4EE to run 3dMark03 at 4.16g on the cpu, 1.775Vcore, then the next it fails at 1.825Vcore and 3.90-3.97g. There is a problem with the multipliers as well, although to be fair, it may be cpu related.....this is an engineering sample with unlocked multies, and I have mixed results at different mutlipliers.

 

for some reason teh EE cpu is an enigma...and doesnt seem to want to play by the same rules as the normal prescotts do. It was the same for the s478 EE's...we had to hardware modify some boards to get them working correctly...and when the next revision came out there were STILL some users with issues on the EE cpu.

 

I would assume that a lot of it has to do with the L3 cache...but I dont really know. All I do know is that the very very few EE users either gloated about how great their setup was, or was pretty pissed that their $900 EE cpu never worked right in this board or that board.

 

Then you add in ES cpu's...I have a 3.4ES s478 cpu and its fantastic...does crazy good clocks and i can play with the multi all day long.

 

some other guys have these ES cpu's and they do weird things...and its mostly that the bios doest recognize them correctly for some reason (even after a bios that specifically updates for ES cpu's), or that adjustable multi somehow messes with the bios's mind lol. I just don't know, and its never anything that we have had consistent feedback about.

 

one guy can build his machine with an EE and have all sorts of issues while his buddy has the EXACT same parts and pieces and the EE works perfectly. Its that oddness that makes it so maddening...and i tried the trick of telling customers "dont use those ES or mobile or any cpu that isnt specifically made for consumer desktop users" and you can imagine where that got me (so i dont do that anymore...i try all the cpu's that were never made for consumers or desktop boards too!).

 

What is weird is that the 16X multi on the board is not that stable for me, while 14, 15 and 17 is doing a good job.

 

more than likely an oddity...i've witnessed a lot of these types of issues with our AMD boards...one multi works only on sundays when you feed it beer and pizza...yet every other multi works perfect with no problems. Its not unusual to see it affect Intel cpu's.

 

 

But, Angry Games said he will chat to the Big Guns in Taiwan.....let's hope they respond sooner rather than later with a very nice Bios update that will sort some of these issues.

 

yep...since i have the board running and i love it A LOT, I am bugging Jacky to bug TW to hurry hurry hurry.

 

but reality has shown (and talking to the engineers themselves proves this) that sometimes what you, the consumer might think should be a simple fix is not such a simple fix...and some of the things that might seem incredibly complex to fix might only take a changed register bit here or there to solve...They work on the bios daily, but they cannot guarantee a time frame on the problem.

 

If the problem was rampant and happened in every test machine regardless of hardware installed and settings used...then yeah, its generally an easy fix.

 

but a problem that rears it's ugly head at certain times but not others, problems that are not consistent all across the board (though they might be consistent across most of the board even)...those can be ones that take a bit of head scratching. Or they can be solved in a simple keystroke.

 

I'll keep annoying them until they hate me (and I'll be starting Mandarin language course @ Boise State University in spring semseter hahaha wont that drive them to the brink) but we'll damn sure get a fix for the 5:4 issue as well as making sure all Intel desktop cpu's work 100% (ie: the 19x 570J cpu's that another customer is having issue with).

 

and I'll keep pushing for a better overclocking bios but you need to realize that there are no guarantees about this ever.

[BigStan]

Nobody is asking for a guarenteed OC. But the advertised specs of a mobo must work. You can't put out a mobo where the 5:4 divider hoses a WIn install or the AGP/PCI lock only works sometimes, and not suffer consequences in taking a hit on the reputation of the mobo. Even the mighty Abit learned this lesson with the IC7Max3. It was pulled from production way before the end of it's useful life because they wouldn't/couldn't fix the VDimm/Vtt issue. And whether you like to hear it or not, DFI is a second tiered mobo maunfacturer. I doubt they can afford to take very many hits on their reputation. The BIOS needs to be fixed.