Heres with 0621 bios....Stuck @ 374FSB:(
http://img99.imageshack.us/img99/935...shot0252te.jpg
32m/win xp/priority...
http://img76.imageshack.us/img76/591...shot0130qn.jpg
Printable View
Heres with 0621 bios....Stuck @ 374FSB:(
http://img99.imageshack.us/img99/935...shot0252te.jpg
32m/win xp/priority...
http://img76.imageshack.us/img76/591...shot0130qn.jpg
Just wondering, have you checked FSB speed with a lower multiplier?
4.1 GHz is a nice overclock though..lol.
Yeah, but imagine a 6600 in there. 380 x 9 is NOT nice.
Maybe FSB will be higher with a lower multiplier though?Quote:
Originally Posted by M1nDH4cK
I'm not really up on Conroe, but isn't 4.1 Ghz pushing it for the frequency of these cores ? Therefore he might get 4.1 Ghz also with a lower multiplier in which case it wouldn't be the mobo and memory holding him back but the CPU itself.
4.1 Ghz....lol, what is that something like a 40% overclock starting with the top of the line chip? I bet the lower CPUs have even more headroom.
4.1 .. no... not really (if we're talking non-aircooling). Imagine having an infinity + conroe e6600 on ln2 for example and not beeing able to go above 400 FSB because of the board... (that's "only" 3.6, and most conroe 6600's surpass that on AIR)
no 12x cpu multiplier ?
My guess is that you will always get less FSB with 11x multiplier than a 9x multiplier especially when you are already 40% overclocked.Quote:
Originally Posted by M1nDH4cK
Kind of reminds me of comparing FX-57 overclocks to opteron 144 overclocks.
Anyway I don't know enough about Conroe yet, but until the multiplier gets lowered I'd have a hard time believing a 6600 would end up with the same FSB limitation.
set PCIe frequency to 101 (or anything other over 100, within reason...). you should get 20-30 more FSB.Quote:
Originally Posted by Dumo
Got new bios...will upload it here:)
interested to hear how this goes
is your mobo modded
http://us.dfi.com.tw/Support/Downloa...FLAG=B&SITE=US
few more beta bios now on dfi site
Nice, DFI made new bios is 5 hours 15 minutes :p: :D
Excellent news, I'm also looking forward to seeing these results. Hopefully they've managed to incorporate similar boot strap/nb settings as the P5W DH to allow for higher fsb without decreased performance.Quote:
Originally Posted by Dumo
Be real intersting to see how far you can push now DUMO mate.
What do all these patches do - dfi says increase conroe multiplier x6 - x50 - does that ean all conroes will have an unlocked multiplier on this board?
Will any of these BIOS make the 975 chipset more Kentsfield friendly?
Cheers
Mav
The Extreme Edition conroes aka the X6800 have unlocked multipliers. Before the BIOS would limit them at the default of the chip (11x) now you can bump up the multiplier.
I'm sure they will release Kentsfield optimizations as necessary.
This is the only thing that is putting me off this mobo.
Lack of FSB.
Is ther any idea if a new bios will rectify this?
All in all is it a stable mobo?
Try it pls with 9x multi, and max FSB , 10x :)Quote:
Originally Posted by Dumo
i'm more than interested to see how this goes dumo
mine also only got FSB 386 :(... BIOS 0621
http://img.photobucket.com/albums/v6...86_dfi975x.jpg
Cal930 in Japan has a Vnb mod that might help some of you. RT9412, pin #6.
He says this mod works well only with 4:3 ratio. Also, I'm not sure I would recommend using 2.0V for 24/7 use. :)
http://cal930.sakura.ne.jp/cgi-bin/c.../dfivnbmod.jpg
M/B: DFI Infinity 975X/G, BIOS: 615
CPU: Conroe E6700 ES
Chipset Cooling: MCX159-CU
Memory: TeamXtreemPC2-6400(DDR2-800)1GB×2
VGA: RADEON X300SE-X/PCI-E
HDD: WesternDigital WD1500ADFD
Power Supply: Zippy-600GE
Vcore: 1.85V, Vdimm: 2.65V, Vio: 3.51V,
Vｎb: 2.0V
OS: Windows Server 2003 Enterprise
http://cal930.sakura.ne.jp/cgi-bin/c...436_11_594.png
Called DFI TW this morning for XE's open multi bios;) Hope it will work.Quote:
Originally Posted by kiwi
This is a nice board with relatively stable vcore and a good range of vdimm voltage options.
i'm still torn between this one and the Asus 975x board :confused:
everone with dfi is stuck on 37x fsb.. hope new bios will help us.. youre not the only one dumo :D
3D with 0707 bios..
http://img155.imageshack.us/img155/9...shot0309on.jpg
lol@Dumo
going OT in his own thread
Just want to see how stable the board with 1.7Vcore and 2.6Vdimm w/out mods:)
Open multis still not work with stepping 5:(
you reckon we'll be able to have open multies for all CPUs :p:Quote:
Originally Posted by Dumo
400FSB possible, but with lower mem. strap...
http://img78.imageshack.us/img78/406...shot0435hi.jpg
Comparo to Asus P5W DH...
http://img250.imageshack.us/img250/7...shot2161ad.jpg
So if you lower the memory clocks you can clock CPU higher and over al is faster in superpi - be intersting to see how big the gap is in 32m :)
Thanks for the update Dumo :)
Generally lower memory clocks will hurt you in 32M. Each of these boards (Bad Axe, P5W DH, and Infinity) will give different results at the same CPU speed. Also, the RAM you use can influence results.
All of the boards I mentioned use the same chipset. But the Infinity seems to have problems with 400+ FSB. I don't think a BIOS is going to fix that. I think it has more to do with the way the chipset was implemented on the board.
Bios 0707 still won't unlock multis upward. It will be nice to run 420X12:)
420FSB is fine if you have an X6800. But if you have an E6600, that only gets you up to 3780MHz. I would certainly want more than that.
I think kyosen said the Infinity had good memory performance. But he and some of the other Japanese moved on to other boards because of the FSB issue. The chipset should be capable of 450+, especially with the voltage options available.
Tony said on BE that the P5W DH may be soft-switching the strap from 1066 to 1333 once a certain FSB is reached (in the background) so the CPU thinks it's still running 1066 strap but the board is running 1333 strap.
This would account for the variances in people's testing results, showing the P5W ahead of boards like the DS3 or 975X/G at the same clocks one minute, and behind the next... just depends on the soft strap it's using. Since at 1333 strap the chipset will loosen latencies giving worse performance at the same fsb.
This is all heresay and conjecture until proof is found, but it's an interesting theory :)
Why revive a 3 week old thread with old news…