Printable View
Sorry but only in Czech language! But graphs dont need a translation, my max OC with on-board cooling was 585MHz full Orthos stable ...
http://www.pctuning.cz/index.php?opt...9222&Itemid=43
My result in English : Overclocking legend is Back, this is best mobo with P35 Express ever made ... brutal overclocking CPUs and RAMs too, great performance ... YOU must to have it!
*First ;)
Yes, first ... some News or pics, and OC result on many web are not a REVIEW!
have u tried not using the coper cpu cover and check difference in temps ?
i just told WWM and i am telling you as well that you both did not comment on stupid placement of RAM slots and the fact longer (gaming) graphics cards will be right next to the slots meaning you will have to take out a GFX card to change RAM
this is a major design flaw in my opinion because every time you have to handle a graphics card take or put RAM in can cause accidental damage...and for people watercooling their cards this would be an additional pain in the ass
http://img394.imageshack.us/img394/9...lpp3526yg0.jpg
http://img89.imageshack.us/img89/382...2e8acb8jg4.jpg
why did you omit this from the article? at least it doesn't seem you've written about it in the negative remarks. Unless i am mistaken and somehow i am not seeing the best angle
well from the look of the top yellow lsot which appears to be openned seems like u will be able to remove the ram without removing the vga, it will be a tight job but it will be done,
anyway i think this is the price of having 3 pci and 4 pci-e slots which i think deserve the hassle to remove the vga ( if true ) every time i wish to change my ram ( which is for 99% of the owners will be very rare )
This problem has many of new boards ... but OK i will update my review with this comment ... THANK YOU!Quote:
Originally Posted by dinos22
oh dear thats not good at all :(Quote:
Originally Posted by dinos22
really?Quote:
Originally Posted by OBR
which other board or boards have the same problem?
For example ... DFI ATI RD600..., Giga P35T-DQ6 + DS4, MSI P35-Platinum,
didn't realise that lol
guess i never really paid too much attention at that board as i had no interest in it........
this is just a pain in the arse problem but one we will all live with nonetheless if the performance is there :D of course heheheheheh
If you want to remove RAMs you have to bend down GFX and it works ...
is that what the user manual says :DQuote:
Originally Posted by OBR
lol. "bend down gfx" :rofl:Quote:
Originally Posted by dinos22
here is nothing to laugh! Look at ma upper GFX! It is bended for good!Quote:
Originally Posted by shimmishim
http://ukazto.com/img/P1020542-4nt1.jpg
"In order to replace RAMS, bend down graphics card, remove the graphics card, or just use a hammer.*"Quote:
Originally Posted by dinos22
*Using a hammer will void warranty.
LOL
looking like a nice board though. :)
go the hammer :hehe:
http://www.pctuning.cz/ilustrace3/ob...etaktovani.png
dis looks good ya?
looks very good adamsleath :yepp: :yepp:
P5K3 as well has clearance problems with 8800 gtx.
I take it the same cpu has been used in these FSB tests? Otherwise it's a bit of a red herring to the capability of the boards...Quote:
Originally Posted by adamsleath
was thinking the exact same thing earlier... my guess is that they did or else you're absolutely right.Quote:
Originally Posted by Johnny Bravo
It is table of capable FSB in day of TEST! Not today! For example ... Striker in testing day (before 5 months) reached only 475, i know today can run at 520MHz, it is explained in text ...
This table is not MAX FSB on that boards NOW, but in day of their first testing! And this DFI won! On first review on our server reaches 585MHz ...
what JB was implying (if i'm wrong, please correct) is that chips have different FSB walls as I have experienced this myself.
A chip will usually run out of FSB before a motherboard does (usually).
So in order to eliminate the possibility it is the chip causing the limitation and not the board... the same chip needs to be used in all tests with the different motherboards.
Yeah that's kinda what I thought OBR - what worries me is some person just looks at the review and assumes that is all each of the board is capable of without consideration for the fact that different cpus will give different top FSB speeds. Still the DFI is strong in that department ;)Quote:
Originally Posted by OBR
OBR,I'm sure that on your next tests some 600 MHz is reachable...:p:
See here.
And try to make your review in English(at least your conclusions-Vyhody+Nevyhody :D ),it's very tricky to translate it from Czech.
Best Regards.
I need to know for a fact guys if the ram slots and graphics slot are going to be a problem for each other primarily if a ram fan is going to be used. I was planning on getting the Corsair Dominator Airflow fan to blow over my dimms and I will be using an 8800 Ultra on this board. Is the long 8800 going to interfere with the ram slots if the Corsair cooler is installed or will I be good to go? I would hate to order all this stuff just to find out it won't work or I have a high risk of shorting something.
On our new web, we will have Czech/English review together ...Quote:
Originally Posted by major_john
Now i am at 610MHz FSB!
Thank you OBR for the test.
Maybe I missed that but did you test for max fsb on all boards with the same cpu? (because every cpu has its own fsb wall)
For your future reviews I would like to see tests like SuperPi (for integer performance) and Cinebench9.5/10.0 (floating point performance).
Performance wise I'm not impressed by the DFI board.
Yes i tested every board with the same E6600, only last DFI with E6850 ... but i think no CPU FSB wall exists ...
To add to this, does it look as if an Ultra-120 Extreme will install ok on this board with the Dominator fan over the dimms? Some of these pics look awfully cluttered in that area making me seriously wonder. I was planning on using 38mm San Ace fans on the Extreme too. I wish more pics from different angles could be used to determine clearance issues. Thanks for the intital write up however OBR.Quote:
Originally Posted by cantankerous
CPU fsb wall exists.Quote:
Originally Posted by OBR
i've tested many chips and many different boards.
two chips will do different fsb's on same board with exact same settings and hardware.
it would be interesting if you could stick that E6600 you used for other reviews into this board.
edit: unless you were trying to say there's no fsb on your 6850.
To clarify 100%.....a chip in the Giga DS4 did 510FSB but the same chip in the DFI did 585FSB+ ?
In your opinion do you think it was down to the board design, or the voltage control the DFI has?
So when is the release date?
I want one.
they've been available for a little bit now (since Wednesday of last week i think?) from motherboardpro.comQuote:
Originally Posted by breakfromyou
check them out. the CS rep told me they should receive more on Wednesday.
Ive tested LOT of cpus and mobos (now i have at home E6600, E6850, E6850 ES, QX6700, X3210, Q6600 ad 5 various mobos) and no one from these has a FSB wall! This si an myth! Every from this chip can work at 600x6, or for example 400x9 ...
I agreed, chips has max frequency wall it is clear, but FSB wall do not exists ... if you are thought you reached FSB wall on some CPU try it on another Mobo, and you will se no more FSB walls ...
For example my Q6600 can reach only 3.2GHz (at any FSB + multi) on 6 various mobos, but when i put it to EVGA A1, it is fully stable at 3.6GHz! It depends on mobo and cpu, no on FSB wall ...
Yeah that's correct. I run my GPU in slot 2 because its so annoying removing it every 5 minutes when I change mem. :down:Quote:
Originally Posted by OBR
Title Fixed.
I have several boards here laid out the same. Easy way to remove ram?Quote:
Originally Posted by dinos22
Roll back the top retention clip, then roll the ram out from top to bottom, the lower clip will roll back quite easily and you can remove the stick. Replacing is simple, roll the lower end in first, once aligned apply steady pressure to both ends of the stick and voila. No need to remove the long graphics card... Works for me every time, I never have to remove a 8800 from the top slot...
regards
Raja
While I agree with your comments on a motherboard limiting the maximum operating speed of a cpu (your example being 3.2GHz -> 3.6GHz) I have witnessed cpu's that have a limit to the speed that their FSB can go no matter what board.Quote:
Originally Posted by OBR
I wonder however are we arguing there same thing here? I am stating that the cpu itself has a limit to how far it's FSB can be pushed, for instance I have tested a x6800 that will stop hard at 510 on a 680i, 2 different P965 board and a P35 board. Other cpu's I've tested have had similar limits cross platform give or take 5-7Mhz.
I have not be inferring that the FSB limit is to do with the motherboard, but rather the cpu itself, and, in relation to the graph I quoted, it is important that the same cpu be used to give an indication of what the boards are capable of.
I agree with you Johnny.Quote:
Originally Posted by Johnny Bravo
However, it appears OBR is stating that CPU's do not have a FSB wall. Therefore, the chip used does not matter.
I.E. if he used chip A to do reviews for the striker and gigabyte, he can use chip B to do his review of the DFI P35 and they should produce the same results.... meaning FSB limitation will be the result of the board and independent of the chip.
My gut instinct tells me I can take 4 chips in the same board and they will all hit a different FSB. I've tried 2 chips before and they hit different FSB's. but again, according to OBR, it's my board that's limiting two different chips to two different FSB walls :shrug:
each CPU has a different FSB wall. Ive had 4...5? E66 and an E67 and they all performed to different FSB, even on the same mobo, and some hit the same FSB ceiling on different mobos. (ie..the chips that I tried on more than 1 mobo)
Johnny and myself have discussed this on a number of occasions. the mobo plays its part, but is absolutely not the only factor
Im fairly certain im not wrong- if FSB is all about the mobo, why does the FSB ceiling increase when you freeze a chip?
I know the DFI will be a good board, but based on your FSB results, I think you used different chips, making the tests skewed...possibly coming across as biased?
Agreed on the cold comment. My E6600 which walls around 510 will get up to 540 on a single stage.Quote:
Originally Posted by K404
@OBR: also, you claim your chips can do 6x600... i'd like to see your QX6700 or X3210 at 6x600. i don't think i've ever seen a quadcore at this FSB before!
Did I claim that? :shrug: certainly didnt mean to- iv never owned a quad*. Also never hit 600FSB- not been above 510 IIRC.Quote:
Originally Posted by shimmishim
*donations gratefully accepted
sorry my comment was directed towards OBR (see up at his post where he claims his chip can do this)Quote:
Originally Posted by K404
The mobo looks very interesting tech/perf wise... but wth were they thinking with that cooling solution...??
Quote:
Originally Posted by K404
Yes you are right about sub zero effects,
sub zero on the board makes it easier for the pwm supply to the cpu, due to decreased trace resistance/hf inductance (lower impedance at hf), giving the processor potential for increased stability under load and higher fsb.
also nb to drive voltage to and receive signal from the processor due to the decreased inductance the voltage amplification and i/v conversion (if there is any) will see, so yes it's part cpu running cold and part motherboard capabilities, as thermal drift and trace layout on the motherboard side to and from the nb will also effect total fsb potential between motherboard brands - based upon the engineer's implementation and component tolerance choices.
regards
Raja
yeah man you'd hope so hey lolQuote:
Originally Posted by Johnny Bravo
but it seems not :shakes:
i strongly disagree with you here man. I have tested :banana::banana::banana::banana:loads of CPUs and there certainly is a FSB wall on CPUsQuote:
Originally Posted by OBR
it is very misleading to post a graph such as that with different CPUs tested;. >> a graph like that has to have ALL OTHER COMPONENTS AND TESTING CONDITIONS EQUAL........i understand that there maybe slight variations in ambient conditions which is acceptable unless the variations are large otherwise your results mean nothing
:hehe:
was the 585 fsb achieved with stock mobo cooling?
im supposing it was.
regardless of cpu's used. i get the impression that this board can effortlessly do fsb speeds exceeding 500 assuming your cpu is capable/not fsb walled.
speaking as an average nob who doesnt have cherry picked stuff :lol:
i spose ill never know whether my old 6700 was fsb limited or the striker board cos both are gorn now...470FSB fer me....i always thought it was the mobo - still do.
and anyway the cpu walls seem to be with the cheapy chips like 4400's/2160's????:shrug: is the e4500 released with a higher multi cos its fsb walled ???
still curious to see x38 fsbs
Actually I don't really like Corsair Dominator Airflow fans too much as they don't really push that much air and high speed small fans = noisy. I prefer to use something like Antec Flexfan or a rigged up quiet case fan to cool RAM slots.Quote:
Originally Posted by cantankerous
I love how everyone is complaining of the PCI-e slot location bent cards etc.
just move the card one slot over... there are three slots...
cpu's have FSB WALL.. :D
are you going to get 16x bandwidth from the other slots????????????????????Quote:
Originally Posted by xlink
So they're going for about $300? I thought they were supposed to be more in the $250 range...:(Quote:
Originally Posted by shimmishim
As far as I know only the top slot is 16x, the others are only 1-4x.Quote:
Originally Posted by dinos22
- There is not one single cap near the cpu!
- One motherboard standoff hole is blocked by the heatpipe going from the SB to the NB.
- There looks to be plenty of room around the motherboard standoff holes, looks clear of wiretraces -- unlike the P965-S.
- There looks to be plenty of room around the cpu retention holes, looks clear of wiretraces.
Great job on the review, excellent qual pictures! It's nice that it's being updated with suggestions here, but it'd be even nicer if I could read it lol. Seriously, I'd love to see a translation, I think it'd be well worth the time.
Spot on about the upper vid card slot and ram. Seems simple enough to swap the slot with the pci-e 1x slot below. I'm no engineer, but it can't be that difficult to swirl the wiretraces a little bit to put the upper vid one slot lower. It's either a major oversight, or far too many technical issues with swapping those slots.
Personally, the green is very garish to me. The DFI trademark black yellow/orange has a higher contrast and more pleasing to the eye under any light. Ever notice that most of the street signs that signify dangerous or important info are always colored black and orange or yellow? It's because the high contrast makes them stick out from the regular signs.
Otherwise, it seems like a stellar board with lots of the DFI trademark innovation! The heatsink at the I/O plate and cluster of voltage regulator components near it and the cpu give it a clean high tech look and should get great cooling with most cases exhausting near it and many cpu coolers able to direct airflow right over it on the way out of the case. Lots of fan headers (6), manual says CPU, Chassis and NB fans are controllable.
This is one I'd like to buy! I'll hold out to see if they release a black/yellow version first. :D
.
CPUs definately have a dramtic effect on FSB. My first E6400 would only do 430mhz fsb, no matter what multiplier, my next e6400 would do 495mhz fsb tested on DS3 and DFI P965, my e6600 wouldn't for 495 fsb on my DFI no matter what multiplier until I put it under cold, at which point it did 533mhz. My latest 3070 does 497mhz fsb on DFI P965 under dry ice no matter if I use the 10x multipler or 6x.
CPU has a very strong effect on FSB these days, even more so than mobo in my tests.
DFI P965 at 1:1? or 4/5 or wich settings?Quote:
Originally Posted by Stapler
edit*
i had this board it's stable, but's its hard to get on speed...
Sorry but the RAM placement isn't a problem. I have the RD600 board which is exactly the same and I can put in and take out RAM with my POV 8800GTX all day long until the cows come home.
All I do is release the lever near the card a small tad, and release the lever opposite and the RAM lifts out. Doesn't make it any harder in my view.
Good point, I've done the same with a few mobos here. The ram usually has not too much trouble getting out when the lock is just far enough open that the keyway is clear. It doesn't have to be all the way open.Quote:
Originally Posted by Mekrel
But I beg to differ in that it's still a PITA in my view -- while the above technique is often preferable to pulling the vid card. On one system recently, a large HSF exasperated the situation and I simply pulled the vid card since it was running open air and easy enough. Far easier to design a board so that it completely solves the issue, and a lot of the latest mobos have the pci-e 1x slot on top and the first vid slot below it. Otherwise, the ram is always too close because it can't move up any higher.
Some people say, "You can't think of everything...", but DFI (and other high end) nearly always does think of everything, and that's what makes it an enthusiast level player.
I sure would like to know if it is indeed an oversight, or simply a too costly technical difficulty to swap the first two slots. It just isn't in character for DFI to do this without good reason.
.
1:1 of course, 4:5 was a crapshoot on P965:D Only person I ever saw do well with 4:5 was coop. 1:1 530fsb on DFI P965 was cake, if you couldn't do that it was probably the cpu holding back the board. Every DFI I saw that was well set up did that kind of FSB without breaking a sweat. God I loved that board, looking through the BIOS the first time on my P5k was so depressing:( Should've ordered the P35 Infinity.Quote:
Originally Posted by BeardyMan