DFI NF4 R.AB0 roll call [revision AB0 OCers post!]

[sauria]

all revisions are exactly the same so far. the rev number just refers to the bios chip it was shipped with. i found this out when i got my ultra d back from rma and it had a new rev number written in with ball point pen, so i called and asked tech support. you can also check this out at dfi street.

-edit- i just saw the above post about rev b**. this may be a newer board.


very truly yours,
politenessman



finally broke 12k

http://service.futuremark.com/compare?3dm05=1210355

That wouldn't explain the variance in Dani's result:

http://www.xtremesystems.org/forums/...9&page=2&pp=25

"My first DFI nF4 Ultra-D had revision R.A02 , did +370HTT easy
My second Ultra-D was R.A03, did'nt try more than 350HTT, was Prime95 stable
Third Ultra-D, R.AB0 did only ~320HTT after many trys (just ones SP1M trough)
Now my SLI-DR with R.AA0 can do at least 350HTT Prime95 for hours"

[Daved+]

all revisions are exactly the same so far. the rev number just refers to the bios chip it was shipped with. i found this out when i got my ultra d back from rma and it had a new rev number written in with ball point pen, so i called and asked tech support. you can also check this out at dfi street.

I can't agree because they do have different chipsets, the new revisions come with the "epoxied" NF4s and the olders don't. I guess they put a new revision number when they started getting different chips from Nvidia.

[politenessman]

i dont know what to tell you. i got my board back, it had the revision crossed out and a new one written in. i posted on dfi street and asked about it, because i was trying to find out if they had sent me back the same board. angry games, the administrator and a dfi employee, and ex roadie, a moderator explained to me about the revision in this thread, if you feel like reading the whole thing.
http://www.dfi-street.com/forum/show...=politenessman



"the only thing changed in the revisions is the bios".
i was also told this by level 2 tech support. as to why this guys something changed this way or that, sorry. i guess these guys(dfi) were lying to me for the hell of it.

very truly yours,
politenessman

[sauria]

Thanks for the link politenessman

[Daved+]

"the only thing changed in the revisions is the bios".
i was also told this by level 2 tech support. as to why this guys something changed this way or that, sorry. i guess these guys(dfi) were lying to me for the hell of it.

I guess "officialy" the only difference between the revisions is the BIOS but unofficially there's more differences. Why they would lie? Because they want people to buy new boards believing they will do as good as the old ones did. But as we all know, in the AMD world, it's not uncommon that the old stuff works better than new stuff (that's why I sold my 0523 week venice and bought a 0517).

[sauria]

"(that's why I sold my 0523 week venice and bought a 0517)."

Know of a place that sells particular steppings?

[Zeus]

So is it safe to assume revision R AB0 means shipped with 3-10 bios?

[Daved+]

"(that's why I sold my 0523 week venice and bought a 0517)."

Know of a place that sells particular steppings?

Nop, sorry, I couldn't find any online either but there was a local guy that was selling one (opened box) so I bought it from him.

[sauria]

So is it safe to assume revision R AB0 means shipped with 3-10 bios?

My R.ABO did

[overcrash86]

no problem wit my new



and fast quick test :

[sirkuz]

1M spi doesn't really say a whole lot for that stability of that high of a HTT...I can crank mine up and run 1M but try running a 32M...thats where you will start stressing the subsystem a bit better.

[sauria]

no problem wit my new



and fast quick test :

It that a R.ABO?

[overcrash86]

It that a R.ABO?

AA0 back from RMA rewriten as AB0

she can go up over 500Mhz of HTT, just need to change mem setting



my first A02 have can bench spi 1M at 515 and screen to 520, then just for fun prime at 475 :





i think just need to set good setting, with good BIOS

[overcrash86]

1M spi doesn't really say a whole lot for that stability of that high of a HTT...I can crank mine up and run 1M but try running a 32M...thats where you will start stressing the subsystem a bit better.

yes but if spi 1M pass at 456Mhz it probably sure to be stable Ã:banana: 400Mhz, without any problems, over 460Mhz, that more hard to stabilize

[sirkuz]

What is the highest you can run a 32M SuperPI?

[overcrash86]

What is the highest you can run a 32M SuperPI?

never tested it, prime stable at 475 it's just for fun, one night

[brschmid]

my .AB0 maxes out at 290ish with stuff in my sig, it could be my ram as well, but i am not going to try lower multis to find out, i have more than enough performance

[dinos22]

my .AB0 maxes out at 290ish with stuff in my sig, it could be my ram as well, but i am not going to try lower multis to find out, i have more than enough performance

that's pretty much your CPU maxing out around 2.9GHz....if you want to max your HTT you have to take CPU and RAM out of the picture

[sirkuz]

never tested it, prime stable at 475 it's just for fun, one night

The next 20 minutes you have free you should give 32m a shot and see what you get. I am curious...I have just starting playing around with my DFI (first time back on AMD since tbird days) and am currently exploring the HTT levels for the cpu and mem sweet spot. I find that prime95 means about squat when looking for my HTT ability compared to large SPI for stability testing in that area.

So far prime95 is for testing the cpu clock max and 32m spi for HTT and memory. Memtest will even pass for me easily while still having 32m spi fail.

[brschmid]

that's pretty much your CPU maxing out around 2.9GHz....if you want to max your HTT you have to take CPU and RAM out of the picture

nope, i was using a 7x multi no matter what i tried i could get 290x7 stable, 290x8 stable (submitted to 2x1gb DB), but not 290x9 because it required more vcore than i was comfortable with using.

currently i am running my computer at 236x10 ram on 233 divider @ 275mhz, 1.35vcore and it is blazing fast

[sauria]

AA0 back from RMA rewriten as AB0

she can go up over 500Mhz of HTT, just need to change mem setting



my first A02 have can bench spi 1M at 515 and screen to 520, then just for fun prime at 475 :





i think just need to set good setting, with good BIOS

Thanks, I'll try your setting, I bet the R.AA0 chipset is better on your board.

[dinos22]

nope, i was using a 7x multi no matter what i tried i could get 290x7 stable, 290x8 stable (submitted to 2x1gb DB), but not 290x9 because it required more vcore than i was comfortable with using.

currently i am running my computer at 236x10 ram on 233 divider @ 275mhz, 1.35vcore and it is blazing fast

it's probably timings or TREF value/drive strength that are holding you back.....or u have the worst luck i've seen.....

[Daved+]

it's probably timings or TREF value/drive strength that are holding you back.....or u have the worst luck i've seen.....

What values do you suggest? I know they're different for each set-up but they may still help...

[sauria]

This R.AB0 is MAX stable now at 312 MHz and no more.

[waiho2k1]

Excuse me, where can I find the R number? I can't find on my board