The RD600 on its own could be good. But then you compare it to what its up against, and you have faster boards with less bugs for less money. This, IMO, makes it a flop.
Printable View
The RD600 on its own could be good. But then you compare it to what its up against, and you have faster boards with less bugs for less money. This, IMO, makes it a flop.
Yeah? So I'm very happy about this and will buy it next days :clap:Quote:
Originally Posted by Pete@X
I see the only user who has problems with DFI is fhpchris and not another. He has gotten a bad board. The rest is happy with DFI, so there are enought comments which speaks for DFI :)
from what i have tested
with amd sistem manager E6600 wont work ....use clockgen instead
cant remember the exact pll but later i will have that
with E4300 the sistem manager works very well.....even the voltages and the ram frequency
dont really know why....:confused:
with a cedar mill sistem manager didnt work either....i had a screen full of stripes and a frezze
thx cadaveca, I had missed that.
don't let the naysayers stop you from enjoying this board, least of all those who don't own it. I have enjoyed it throughly, I have not had one board from any manufac without issues. In fact visiting support forums for all major manufacs one can see 'issues' on a par with these.
It was the users themselves that hyped the board like lunatics, not the test team. A grand case for tight moderation of pre release I want it to do 600fsb threads.
2nd that :toast:Quote:
Originally Posted by raju
@fhpchis:
If you want to kill a board - there are always possibilties to do so - so that fact does not count. IF you run high VCores without the needed precautions (cooling) - you are the one to blame - not the board :nono: ! Running high Vcores on the Board without the needed steps regarding cooling is a typical case of the "user.exe needs an update" :nono: .
You mix apples with onions with you "bad" DFI list, why not join the folks on the "680i is killing ram" thread ??? ... i have tested ~ 2000 CPUs on DFI nF4 Boards (part of my Job :D ) - not 1 cpu died on my Hands, Expert Issues where mostly with pre-rel. boards or alpha bioses. So the useres did know that something could go wrong - oc needs also some brain to work - just get stuff, cranck up the voltages has not always the best effect. Who really cares about an E6600/X3060 @3.7 or 3.8 with a VCore passt 1,5V ??? No1 to tell you the truth - far more interessting are the nice results in the low Vcore regions. These results leave enough room to tweak the system for stability and performance.
You call the Abit a good Board ? I could get easy ~ 20 customers who would tell you the opposite - would that be a reason to judge the Abit a "crappy" board - i my eyes no - as there are always two sides of the coin. Some users can handel a board better than others - simple like that.
VCore - why that high vcore (1,67 and so on ...) ? Ever had the idea to find better cores that needed less juice to perform nicely ? There are dam nice Xeons around ... 3.6GHz ~ 1,3V, 3.8GHz ~ 1,4V - there is absolutly no need for insane VCores. I have now 5 DFIs up and running, all @3,6GHz or 3,7GHz below 1,4V. More than enough for 24/7 use :D - but i guess useless to say that to you :rolleyes: .
Anyway - i have the 6th DFI here, running P 25.1 since +12h a X3060@3.7GHz@1,375VCore. PWM HS is easy to touch - no burn marks etc .... ;) - so before i extend that post - i`ll go better back to my work and skip the other thoughts i had in mind ...
If someone whats to buy the board, has the right feeling about what to do - and what to skip - very well, the board will make you happy.
Just a final note - since i play with boards (+15 years) - not one has ever died in my hands / nor rams or so .... we always have to keep in mind > treat the hardware nicely - and you have fun. Go over the limts and you are lost. Simple as ABC - if you follow these rules, all will be well :D
For my side - the case is closed - as there will be always 2 sides, and i`m on the "nice board" side ... so why bother, you ruined a board in a short time. Sorry for the board - not for you - as mentioned above.
Interesting to read, do not have this board -but a nice reading and a nice experience how different people behave themselves in the discussion :)
:clap: kudos Pete
PeteX...would you say 1.525v on a e6600 is too much??
Show me 500fsb with quad core plz :D
Honestly DFI Infinity 965 is the board to get as long as you don't need SLI/CF
Quote:
Originally Posted by kiwi
I second that. :)
Quote:
Originally Posted by Pete@X
Hey Pete,
I have two points to make here:
1. This is a $250 board, much more expensive then most core2 boards. one of the features is a high vcore, if high vcore kills the board, then what good is it? Saying you don't need it doesnt matter, its part of what you pay for, and i for one expect it to work. HOWEVER, i do agree that FHP's case is isolated and he probably just got a bad board.
2. My second point is this, even without FHP's issue, this is still a $250 board that can be outperformed by pretty much any other board out there. It's clock for clock performance is simply awful. I only hope that the new beta bios bingo is talking about really does improve sync 1t speeds as it seems that is the only way this board will be able to compete clock for clock with the other chipset.
sync 1t is the way to go for those with sub ddr1000 capable ram, for those who have the good stuff, ddr1100 and over at cas 4 2t outperforms 1t cas 4 ddr975, on current bios.
The whole point is that I DONT want to kill another board, but I am sure that if I put this thing through a fair second test, It will die.Quote:
Originally Posted by Pete@X
Dont babble at me saying anything about the voltage that I am running is bad/scares you/degrades/whatever. I clearly have enough cooling here, as I am the one running 8+ hours prime on this voltage...Quote:
IF you run high VCores without the needed precautions (cooling) - you are the one to blame - not the board :nono: ! Running high Vcores on the Board without the needed steps regarding cooling is a typical case of the "user.exe needs an update" :nono: .
I also have run 1.65v+ daily, so your talking to the wrong person here.
WOAH! BECAUSE MY 680I DOES NOT KILL RAM! HOW DIFFICULT TO UNDERSTAND!Quote:
You mix apples with onions with you "bad" DFI list, why not join the folks on the "680i is killing ram" thread ???
Is that what Kyle @ [N]ubocp told you?Quote:
... i have tested ~ 2000 CPUs on DFI nF4 Boards (part of my Job :D ) - not 1 cpu died on my Hands, Expert Issues where mostly with pre-rel. boards or alpha bioses. So the useres did know that something could go wrong - oc needs also some brain to work - just get stuff, cranck up the voltages has not always the best effect.
I donno? POSSIBLY XTREMESYSTEMS.org?Quote:
Who really cares about an E6600/X3060 @3.7 or 3.8 with a VCore passt 1,5V ???
Who cares about low voltage? Why dont you just run stock? I booted with my X6800 for two days @ 6*435 and I didnt really notice in day to day things, whats the point? Again I say over and over, your 3.7 isnt holding a candle to the 4100 I am running 3dmark at on water, and for those things ~500 mhz and a faster chipset is kinda noticeable...Quote:
No1 to tell you the truth - far more interessting are the nice results in the low Vcore regions. These results leave enough room to tweak the system for stability and performance.
And again, I ran 3.9ghz on 1.54/1.6+/whatever I want, and it will pass any stress test that I throw at it. I am the one talking to real people here.
I am the one here, helping others like Eva2000 get those extra 50-150 mhz out of his X6800, all by using the same settings that ive talked about over and over and over and over and...
Quote:
Originally Posted by EVA
1) I bet over half of those people have E6300/E6400sQuote:
You call the Abit a good Board ? I could get easy ~ 20 customers who would tell you the opposite - would that be a reason to judge the Abit a "crappy" board - i my eyes no - as there are always two sides of the coin. Some users can handel a board better than others - simple like that.
2) I bet over half have cheap ram (under 300$/2gb, not counting corsair)
3) I bet over half those users run the stock aircooled chipset
4) I bet none of those users have modded boards
5) I bet only 1% of those users even knows what a digital multimeter even is
Again, the AW9D isnt a noob-friendly board, but apparently neither is the RD600. Hell, Kyle@[T]ardforums only managed 333FSB on the AW9D. How sad is that...
Because it is fast. Duh. XS = FAST(benching or daily)Quote:
VCore - why that high vcore (1,67 and so on ...) ?
Yes, I called intel and asked for a 3.9-4ghz cpu @ 1.3v and they said call newegg.Quote:
Ever had the idea to find better cores that needed less juice to perform nicely ?
:stick:
3.6 is slow as crap, 3.7 is below average, 3.8 is average, and 3.9 isnt too bad. Oh yea, all those are 8+ hour prime.Quote:
There are dam nice Xeons around ... 3.6GHz ~ 1,3V, 3.8GHz ~ 1,4V - there is absolutly no need for insane VCores.
XS is about pushing the limits of computing, daily or otherwise, and I think my 8+ hour orthos @ 3900+ mhz kinda counts on that arena.Quote:
I have now 5 DFIs up and running, all @3,6GHz or 3,7GHz below 1,4V. More than enough for 24/7 use :D - but i guess useless to say that to you :rolleyes: .
3.7 is slow. My L623A did 3.7, and people like you couldnt even get 3.2 on the same stepping L623A E6600.Quote:
Anyway - i have the 6th DFI here, running P 25.1 since +12h a X3060@3.7GHz@1,375VCore. PWM HS is easy to touch - no burn marks etc .... ;) - so before i extend that post - i`ll go better back to my work and skip the other thoughts i had in mind ...
I am sorry the board is a POS as well, back to the store it goes...Quote:
Just a final note - since i play with boards (+15 years) - not one has ever died in my hands / nor rams or so .... we always have to keep in mind > treat the hardware nicely - and you have fun. Go over the limts and you are lost. Simple as ABC - if you follow these rules, all will be well :D
For my side - the case is closed - as there will be always 2 sides, and i`m on the "nice board" side ... so why bother, you ruined a board in a short time. Sorry for the board - not for you - as mentioned above.
Sub DDR 1000 capable ram cant do sync 1t ;)Quote:
Originally Posted by raju
i dont care if at one specific setting it beats 975 it is still slower then 975. You even proved it with your personal benches.
SO, like i said if the board really does have bios fixes for 1t and sync clocking that is its only real chance to become a good core2 board, as that is the only time clock for clock performance is good. If we can do 450 1t 3-3-3 sync this board would kick ass. But we can't. ;)
Fact of the matter is this, people will like the board regardless of whether or not it lives up to your personal requirements.
If you have 1000 capable ram and you buy this board to overclock run 2t with overvolt ddr1100, no brainer. No need for dodgy winks, sub 1000 ram can be overclocked, or are we talking stock settings on a dfi lp board now?
Everyone knows the 965 is slower in nearly all current forms than the 975 board, but does that make the 965 a failure? Unrealistic expectations is what springs to my mind...
There is no way I believe that chris gave his bioard a harder hammering than me, he did not even come close to what I've put through this board, so that to points at him being unlucky..
I dont know why you are so defensive about this, i am not telling people to dislike thier boards. What i am saying this is a $250 baord that underperforms. Certainly it works ok, but then compared to its competition it doesnt.Quote:
Originally Posted by raju
Certainly 965 is slower clock for clock then 975 - but then it doenst cost as much either. I payed $180 for my P5B-deluxe, it can perform similar clock for clock on RD600, and it can hit higher FSB. Fact is RD600 should have been infinity, because this board is supposed to be top of the line, but its not, its just top of the line price.
Im not hating on DFI either, i have many of the same gripes about most asus boards...
Also my point about memory performance is that 2t 1100 is not fast on this board... i dont why its so hard for you to grasp, as you proved it with your own tests. for the last time, the only time this board competes clock for clock is in sync. BUT sync mode doesnt work - either from a chipset limit or a bios limit fact is it isnt there. Once again the only hope for this board to be a real performer is in improved sync mode clocking. Either that or they make the async timings tighter.
Otherwise, ill pass on a $250 board that doesnt do any better then my $180one , thanks. async mode doesnt interest me - what good is it if performance isnt improved. i dont care if my ram run 300mhz or 1400mhz all that matters is its performance (See: latency) Once agian, so you don't get your panties in a wad again, i am not saying its a bad board, i am simply saying you dont get what you pay for, therefore IMO it is a flop.
If people want to go on paying too much for a board and liking it - as you say, they are welcome to it, regardless of my personal requirments(which are only that a board offers good value for what you pay for).
My 2 cents to this.
iCFX has the tweaks to make conroe and kents clock high, it also features high end caps and other components that drive the cost of the board high. All you guys want boards that clock CPU's without the need for mods etc but most do not want to pay for them...kinda sad if you ask me.
let me give you an example:
A while ago I priced up manufacturing cost for the reference RD480 board from ATI, at the time it would have been the best Xfire board on the market. Without packaging, CD, cables etc etc the board was $174 just to make, add in all the other stuff you need and you were talking 200+ INC packaging shipped to the USA. Now i know you would have all freaked out so the project died.
My point here is you really do need to get a grasp of how much these boards cost to make...you want the best caps etc you have to pay for them, you want the Wu creating boards that clock like mad..you have to pay for his RnD time.
Simple facts....if you want to buy cheap boards there are plenty out there, if you want the boards that clock well then they always cost more.
hell the i680chipsets were 100bucks at launch with nothing else...so the board was costing 100dollars without even a PCB.
Quote:
Originally Posted by Revv23
It is not hard for me to grasp, given that i actually have the board and know the figures thus far either way in my own comparisons on my system.
What matters to users who actually own the board at present is where the speed is at for them personally at this time, it just so happens that ddr1100 cas4 outperforms all other possible settings below it on the current bios, which is what I mention it, not to satisfy 975/965 comparisons, I know that thru my actual testing of various sysnc/unsync 1t/2t/3t testing, there is no shame in telling that to users of the board, that's what I am defensive about, the fact that poeple are pouncing upon those who have the board with the glories of other chipsets all the time, at least let them get the best perfomance thay can from their own board, which does not have to be the fastest in the world.
The only place we actually agree is that bios work is needed on sync, but from what I know already the gap between sync ad the speeds I have hit up high in terms of benchable latency bandwidth (ddr1226 4-4-4-8 trc 8) is so large that sync is gonna need to come upto around ddr1000 or so to beat it for me personally for everyone else who owns the board they will have to find their own lower sweetspots, and at this point I'm not talking 975/965 comparisons.
thank you for your input.
raja
When is price an issue if a piece of hardware is the top dog?......it's not.
Only when performance is about equal is price an issue.
Raju
The bios in 1:1 sync mode uses the 1066 ram table which is the most agressive, this is why it is a lot harder to overclock in this mode.
Also the chipset hits its max clock at 464fsb, there is no point trying to clock higher than this really as the memory performance takes a dive. This is why i kept saying buy CPU's with higher multi's for this board.
T
this weekend we benched a couple of rigs for max 'pi on air with as much voltage as still improved final speed.
the only two components switched between the bench sessions were an RD600 and a p5b-dlx.
even the OS on the drive was the same.
after several hours using an e6600 at around 3.7ghz we could not get iether board to pull out a convincing lead.
now we maxed at around 427fsb with both boards (RD600 went a bit further possibly due to finer vcore tweakability)
luckily we could use the 4:5 divider on the asus giving us +-533mhz for the D9
we also ran the Rd600 memory at the 533mhz mark.
the final score was in the favour of the p5b to about 0.06 seconds...
BUT we had not yet used the RD600's ability to run the ram way higher with as much voltage as we could want really.
in the end the flexibility of the RD600 would have led to it taking the super-pi 1M crown had we taken the time to push the ram further.
the p5b was maxing on the cpu and further ram speed was therefor out of the question.
so once the p5b has gone over to 1333strap even an upwards divider cannot get it to pull out a lead on the RD600 and this in 'pi 1M supposedly RD600's weakest link.
in the end this board is so much more flexible, requires no vmods, looks great and performs equally to the most popular mobo out there.
it definately has its place in the sun
plus who wouldn't want to run the mobo with AMD chipset for intel cpu's :D
The facts are this, The board clocks well, specially with a kents, the issue is its not as fast.Quote:
Originally Posted by dogsx2
now is that the boards fault or the chipsets fault? I think the chipset is at fault a little here. Now going by what we all have seen you will either live with the chipset limitations and clock around them or you buy a board featuring another chipset that is faster.
Thanks for all of your input Tony. Can you go into more detail about sync mode? Will there be more bioses available for this board in the future, to your knowledge?Quote:
Originally Posted by Tony
Thanks :)
Quote:
Originally Posted by Tony
Yeah I know Tony that's why I only use an x6800 with this board, however for outright benching in 3d marks you can get lower latency and higher bandwidth benching scores by nearly killing your ram lol and murdering the chipset... Once you get over 4ghz it's even harder to get sync speeds to co-operate, sub 4.2ghz you can use many of the post 400fsb settings, for quick benches upto say 429. At 11x multi the 266 strap is good for upto 425fsb in async and can be forced to bench at ddr1200 cas 4, leaving sync on the current bios (with it's inherent limits) for dead.
Thus far highest becnhing perfomance for me is fsb511 (max in bios), ddr1226 cas 4-4-3-4-8 trc 8 trfc 34, she will pass aquamark with that and run 3dmark2001 at ddr1208 at the same timings, read data transfer of 5, trrd 4( can go 3 but actually worsens performance), clocking drive strength needs to be at 13-14, however the core users are best of with the 266 strap and sub fsb you mention for 24/7 use.
:fact:Quote:
Originally Posted by Tony
:woot:
LoL. DFI is NOT to blame.
:toast:
I agree with tony and dogs..
I never thought DFI has done a bad job, the chipset doesnt compete... so i stay away... thats why, as i said, i think DFI's RD600 would have been a better effort as an infinity series... there is no need to make a top end board out of a bottom end chipset.
like dogs said, the price of the board can be high as long as performance is there.