[moved to BIOS Factory section] NF4 5/10 Beta BIOS

[Guest eva2000]

Well 510-3 was fine with my 4000+ sandy with 1GB Gskill PC4400LE at low multipliers i.e. 12x234HTT 2808mhz at 1.67v but soon as i went to 10x280mhz 2.5-3-3-7 1T i corrupted my OS and had to do a winxp pro sp2 fresh install. But with 310 official bios it handled exact same 10x280mhz 2.5-3-3-7 1T no probs.

 

flashing back to 310 official as with 510-3 my temp sensors went all wacky showing 128C on all sensors but chipset heatsink wasn't even warm to touch lol

[PGBK]

winnie 3200+ & TCCD w 510-1 on orange slot => loose some Mhz

winnie 3200+ & UTT w 510-1 on yellow slot => loose some Mhz too

winnie 3200+ & UTT w 510-1 on orange slot => gain 1 Mhz

 

yellow slot and orange slot does make a different but did not bring my UTT higher w 3.24v and need strong drive strength but can only manage 245Mhz

 

orange slot w UTT and weak drive strength of 1 bring my UTT to 248Mhz on 3.24v

 

orange slot w TCCD loose 2 Mhz both for 2.5-3-3-7 and 2.5-4-3-7

[johnrr6]

Well 510-3 was fine with my 4000+ sandy with 1GB Gskill PC4400LE at low multipliers i.e. 12x234HTT 2808mhz at 1.67v but soon as i went to 10x280mhz 2.5-3-3-7 1T i corrupted my OS and had to do a winxp pro sp2 fresh install. But with 310 official bios it handled exact same 10x280mhz 2.5-3-3-7 1T no probs.

 

flashing back to 310 official as with 510-3 my temp sensors went all wacky showing 128C on all sensors but chipset heatsink wasn't even warm to touch lol

 

Eva----I have the exact same batch 4000+ Sandy as you....

 

I'm starting to come to the conclusion that this version CPU just does best at default (12) multi I do best at FSB 234-235 same as you....

 

doesn't seem to like Multi 11 or 10 at all

 

Doesn't seem to matter what bios......

 

Doesn't seem to matter what slots---yellow or orange.

 

The good news is that I can run my BH-5 VERY tight at that FSB with around 3.3 VDIMM----and it runs cooler, quieter, and the bench scores are great.

 

The memory benches are somewhat offset by the very tight timings----and the gains in stability----especially game stability seem excellent.

 

Maybe not the high FSB everyone is so enamored with-----but she runs like the razors.....STABLE.......John

[jikdoc]

i have SD3700 0515 spmw

same exact issue with fsb limited at 233-5.

because of 11x multi, i need to use dividers with my utt. only 5/6 divider will work at 288+, with mem ~ 228. 9/10 divider is a no go, tried many different bios, changing ldt/chipset/vcore/vdimm etc. i noticed however, that i seem to be limited with this setup by fsb at 300. not sure if this is a coincidence, but 300fsb puts my memory at 236 with 5/6 divider.

i have the same experience as johnrr though, using tight timings gives me bench scores around the same as my fx55 which ran utt up to 250 with tight timings at a slightly lower overclock.

[Radical_53]

Maybe not the high FSB everyone is so enamored with-----but she runs like the razors.....STABLE.......John

 

... which I never understood, as the high HTT never gave me a "real" perfomance boost. I'd still be running my OCZ 3500 EL Plat if they hadn't given up half a year ago, got the TCCD for an exchange

[soundx98]

SD 4000 0515VPMM -

 

have to agree with you on my chip liking lower multipliers with this bios (5/10-2).

I did manage to get it OCCT/prime stable at 280x10 with a 180 divider. 282 causes Prime to reboot system after 5 minutes.

 

Glad to hear eva2000 got 280x10 back with 3/10 Official. (Dang my 3500 Winchesters liked that bios, I really need to give it a try with the SD 4000.)

 

I was really hoping to see close to 3G with the 4000 and this bios. Man am I naive.

but it's really running well and honestly other than benches, the 235x12 and 250x11 configurations feel almost the same in real world stuff. Don't know if it is the SanDiego core, 5/10 bios temps, or the Shin Etsu but this runs substantially cooler than the winchesters did.

 

johnrr6 - just want to throw out my congratulations on the excellant memory compilation you have done. Big thumbs up. I know it was a lot of work.

[billy1ear]

I'm seeing both 1 and 3 for Tccd....which one is best to use?

 

No one seems to want to answer this question, I have been following this post, maybe I missed it. Many seem to know, perhaps someone would share when you use -1 -2 or -3. Should be mentioned in the first post I think.

[century child]

"You will get differing opinions on the BIOS , but here is mine:

 

Any -3 BIOS version will be the easiest to set up with TCCD.

 

The -2 version has tested very well with TCCD and UTT (Winbond chips), but must be tweaked a bit more. Angry says this one is the best overall." - AndyOCZ

 

This was Andy's response in another thread to the same question.

[RGone]

et al. No one is doing any bios recommending by variation for many reasons; paramount of which is > each board has preferences not common to another as seen by most users.

 

The first thing to know and understand is in this link...

 

The possible which version is contained in this link... and is loosely based on what or how your setup acted previously and what configuration you were able to use before.

 

Now those of us who fool with bioses all the time have learned not to do much recommending because of such widely experienced, variances. Especially not in an open forum where every single suggestion is met with why/what/ or oh my oh my > mine is failed or did not boot or I don't know what to do next.

 

I have already seen too many that have failed to boot up entirely when they flashed and had to go to some supersaver memory just to boot again. Therefore it really is not up to those without ways to recover and skills to recover to flash these beta bioses > see first link above.

 

Be frikken careful if you do not know what is what. It really is a truth and will keep you from harm or pulled hair.

 

RGone...

[billy1ear]

"You will get differing opinions on the BIOS , but here is mine:

 

Any -3 BIOS version will be the easiest to set up with TCCD.

 

The -2 version has tested very well with TCCD and UTT (Winbond chips), but must be tweaked a bit more. Angry says this one is the best overall." - AndyOCZ

 

This was Andy's response in another thread to the same question.

 

 

Thanks,

Just found the same post and was about to quote it but you saved me the trouble.

Thinking about flashing this bios, looking for that "warm and fuzzy feeling" I guess.

 

After RGone's doom and gloom speach, I think I'll watch this thread for a few more days before I jump in.

Thanks again for the quick replies.

[SHiRaKaWa]

@ Angry_Games

 

I know this is out of topic. But I couldn't help it.

From where did you get your 4400+? :drool:

[RGone]

After RGone's doom and gloom speach

No sir that is not a doom and gloom speach nor post. It is a reality check. It is the truth and not anything anymore.

 

Another truth is I run any bios I want with any of 6 or 7 types of memory stick. See what I wrote. Yep, that is another truth. I have to go way way out of the way to try and duplicate these failures many users see. I have to go way out of bounds to get problems and then sometimes it takes hours and hours to duplicate a failure similar to some users failures. And then there are times, I never see the failure ever. Cannot get there with my parts and pieces.

 

Maybe such as this is why. I toured all the other motherboard makers forums before DFI "ever even released" their NF4 boards. Particular attention to the users in those forums that made a whit of sense> it quickly became apparent that native 24 pin power supplies were going to be a near necessity. So what happened? AG and I had native 24 pin power supplies at the first crack of the starter's gun for DFI NF4. Saved us a ton of headaches.

 

Then from research with the NF3 A64 boards it was understood that A64 did not need super bandwidth generated by 2,2,2,2,2,2,2 types of timings and that meant TCCD based memory was going to run more easily on the NF4 boards in what is Dual Channel and not single channel as was the case with the 754 boards. So prior search had many of the problems sorted before AG or I even saw the boards.

 

It is the basics that trip most folks up. The unsure knowledge between rights vs privileges as is the theme of overclocking for those of us starting a few years ago before mother board makers jumped on the OC bandwagon and supplied added tools to 'enhance' but not guarantee any overclocking.

 

Oh yes longwinded. Not doom and gloom but every now and then a re-focus and reality check should come into view for understanding and reassessment.

 

RGone...