looking good guys
anyone able to show some 340MHz+ HTT stable OCs
Printable View
looking good guys
anyone able to show some 340MHz+ HTT stable OCs
up to 3.6V :confused: i thought this mobo goes up to 4vQuote:
Originally Posted by Dumo
if the mobo goes up 2 4v, does this mean that you can up the 3.3v rail, and get volts over 4v?
You can if the memory needs to run @ 4V :)Quote:
Originally Posted by dinos22
http://img475.imageshack.us/img475/466/mvc148f4lt.jpg
Hound no it actually runs off the 5v railQuote:
Originally Posted by Hound
Any luck with the new bios?Quote:
Originally Posted by MickeyMouse
can someone do a direct compare with their nf4, i.e. your best with the nf4, and then your best with the rdx? i'd like to see what speed we need with this board and timings to catch a 640+ HTT nf4.
thanks guys
Guys
Both Robert(Rgone) and myself have come to the same conclusion with this motherboard.We based our decission on the following info.
I have been speaking to Oskar at length regards 300+fsb, he is trying his hardest to get the board clocking this high but we have decided it is just NOT WORTH IT
We have found the board to be a stella CPU overclocker, it clocks cpu's higher and usually with less vcore and less heat than the NF4.
So, forget about running DDR600, stick to BH5 or TCCD with 2.5-3-3- timings and push with a higher multiplier, we have a feeling you will get faster than you did with the older NF4.
Rgone has forwarded benches to me with a crappy venice he has NEVER been able to reproduce on other boards...280x10 or so with 2.5-3-3- but the beauty was the cpu was running less vcore, a higher multi and a higher final clock speed...all prime stable.
So, after beating on this board we gave up with 300+fsb as being honest this board does not need it, enjoy a higher CPU clock and just tweak the memory in just below it.
Tony
does the ati or nvidia board perform better clock for clock?
there seems to be some issues, two people had it. Might be powersupply related.http://www.xtremesystems.org/forums/...ad.php?t=77607
With a divider...I think (not 100%) he is referring to clocking 1:1 at 300....I don't think he meant the board won't run 300+ HTT.
Quote:
Originally Posted by Tony
what if i need 315 mhz to get the maxout of my cpu and memory?
what if i pay 230 euro's for a motherboard thats clocking lower then my MSI K8N neo2 or a DFI NF4
saying the board makes the coldbug history, and I CANT get this board started with a 500W 24PINS native psu!
This is my first own experience with DFI and its the worst!
Maybe MSI doesn't deliver ALL the power dfi can deliver BUT they boot with every psu!
:brick: :brick: :brick:
It's really a pain. I just went to your thread at DFI streets and people are recommending that u use an 8pin. So thats like a $300 upgrade with a new PSU. Even if some of us send back out PCPC to get it modified, it still costs money and shipping those bricks arent cheap. THis might be the only consumer board with an 8pin. So whens BT coming out based on the 4pin Halibut reference board?
helow situman, i feel VERY angry about this situation. dfi saids to buy a psu thats in there recommondation list.
is there no possiblity TO even try to get all psu's working and RELEASE a mobo that has worked the biggest bugs out of it.
http://www.dfi-street.com/forum/showthread.php?t=10854
I didn't know about these facts.
Quote:
if the board has a 24-pin power connector, and we state (as we have) that you need a true 24-pin power supply and that a 20-to-24 pin adapter does not work and is not supported, then you need a true native 24-pin power supply.
Quote:
we tell you that you are required to have a 24-pin 480w power supply because that is what is necessary.
Quote:
i'm tired of some of you who think you know more than our engineers continuing to argue that we have no clue what we are talking about and you can skirt the minimum requirements.
i have a psu that has these specs. and i have only 1 thing To say. BETA
Run a divider to clock the cpu high and the memory a little lower, 300 2.5-4-3-8 can be slower than 280 2.5-3-3- so always run tests to make sure what is fast and what isn't, sandra bandwidth is NOT a sure way to prove you have a fast system, unbuffered sandra is closer but ulimately you need to run real life apps and benches to be sure. I ALWAYS use lobby Low and drago low in 3dmark 01 to test ram, latency and overall speed...its a real good indicator to what is fast and what is not.Quote:
Originally Posted by Jort
Also please remember just because you have issues does not mean Joe Sixpack next door will have the exact same issue.The board is a lot different yes, regarding psu's i know the powerstream runs the board ok and Rgone has used PCP&C as well as another dedicated 600W unit so good high end units seem to run it just fine.
Every board has issues, MSI has the 9nm issue with the ealrly boards, asus had real poor overclocking on the early boards and yes there may be an issue with psu's and this RDX200 but its too early to bash the board as poor and not worth someones time.
run the board 9x300 with the ram in 183 mode, CPU MHZ are more important than ram ultimately but keeping the ram latency low will keep speed up.Quote:
Originally Posted by MickeyMouse
Remember there is no added latency running a divider on A64, you don't have to run 1:1 to make sure its fastest.
......please answer my question
Quote:
Originally Posted by Tony
hey tony,
when you buy a car for ********, you would like to have it runs smooth don't you?
i can run my mem @ 314 2.5-3-3-7 so this is what i like to do.
i understand this board is just released but that it is so picky about a damn psu?
seems i can't change anything about that :(
The board to me is clearly a beta board after more testing.
The chipset is fast but should of been tested more or the uli chipset should of been used.
My board runs hot but I guess thats the nature of things.
I totally understand that but when a board is promised to do such and such we tend to expect it.Quote:
Originally Posted by Tony
This board has really been a letdown in my opinion.
But then again i am just a customer
Having the reputation dfi has in the overclocking world, how can they release a board this way? We are talking about a board that is supposed to be the very best in the group of the best, to be used by the best overclockers out there, what are they thinking about? people doesn't pay that amount of money to get a mobo out of the box and find all these problems.
The 1575 came available last week....so you are saying DFI should have redesigned the board and gone with that chipset with the board coming in 2 to 3 months?Quote:
Originally Posted by althes
ATI have a new chipset coming in that time period so maybe DFI will opt to use this with the 1575....
One last thing you all need to remember Sb450 is not pin compatible with ULI1575. If you are thinking how can abit and msi etc all come with 1575 so soon its because they started with 1573 then just dropped the new chipset in place.
Heres a more fitting analogy, take a turbo charged car and turn up the boost with a weak fuel system and see what happens (car is to motherboard, as fuel system is to_____?) :p:Quote:
Originally Posted by Jort
sorry, i couldn't resist your car analogy. i'm sure your fustrated, and i hope my experiences aren't the same :)
The issue is you are comparing to the NF4, that was tuned balls out to run TCCD at super high clocks, this board was not.Quote:
Originally Posted by althes
Its taken me 7 full days to realise this board does have promise, it took me 6 weeks to get halibut running well and many bios files.
There is time yet for this one also, but in the world of the enthusiast you all want it to run awesome from the getgo, and im afraid that is real hard to do.
70 bios files on the NF4 ultra D over 10 months prove this, and yes i have access to every single one of them.
DFI will drop in SB600 early next year, because it's pin-compatible with sb450. That's what I read at anandtech, and makes sense to me.