Printable View
Well this is all I can squeeze out of this combination...
vcore=1.95, vdimm=3.3, vdd=1.9, agp=1.7
timings= 11-2-2-2-2-9-12, enabled, enabled, fast
cpu interface=aggressive
http://www.creaseyprinting.com/scree...x10SuperPI.gif
http://www.creaseyprinting.com/scree...x10SandMem.gif
http://www.creaseyprinting.com/scree...ndCPUArith.gif
http://www.creaseyprinting.com/screens/250x103DMark.gif
Sorry for the large images....
:banana: :banana: :banana: :banana:
I am very happy with this set up!!!
really a good achievement m8! Just a couple of questions:
1) Is vdd 1.9v stable? I noticed on my board that going over 1.7v makes things act worse.
2) Is your ram single or double sided?
3) Does ddr slot3 (the one on the right side of the mobo) overclock less than slot 1&2?
Very Nice:)
Thanx....Quote:
Originally posted by Iolao
really a good achievement m8! Just a couple of questions:
1) Is vdd 1.9v stable? I noticed on my board that going over 1.7v makes things act worse.
2) Is your ram single or double sided?
3) Does ddr slot3 (the one on the right side of the mobo) overclock less than slot 1&2?
1) It seems to need all the vdd
2) Double sided (I think) it is OCZ PC3200 rev2, bought Jan03
3) Not sure... I did all this dual channel slots 2 & 3
Have you some sort of cooling on the mosfets? If you have spare time can you test a single stick in all three slots?Quote:
Originally posted by clockedOut
Thanx....
1) It seems to need all the vdd
2) Double sided (I think) it is OCZ PC3200 rev2, bought Jan03
3) Not sure... I did all this dual channel slots 2 & 3
Stock cooling on the mobo.
I did try single stick with my locked Barton... but it didnt make any difference... I was maxed out on the CPU at 225x11.
I will try moving single sticks around later.
Very nice, and certainly very intriguing as we've seen such mixed results from this board and the LPB.
Was 250fsb prime stable ? Where did you get the board as Im looking for a replacement for my 8rda+ and the abit boards dont seem to be doing so well right now.
Wow not too bad!
GRRRRR hurry up excaliber!! I WANT MY INFINITY!!
No, 250x10 is NOT Prime stable.... I have to back off to 245x10 to get Prime stable at same volts.
I bought it from NewEgg 2 weeks ago.
very nice clockedout excellent job Im gonna be ordering one of these board sometime this week for me and a friend.
Very nice ;)
This will be my next mobo for sure
:banana: I like this mobo!!! :banana:
I still think it has voltage regulation issues.
And I wish it would let me use vcore past 1.95 or didnt undervolt.
If I could combine the stability of my Soltek 75FRN2-L with the votage options of this Infinity... it would be the perfect clocker for me. :D
It sounds like anice board but no way no how can I live with 1.95 Vcore. I just couldn't do it. Even 2v even is no good for me. I want and need at least 2.1 v
No one has tried a vcore mod for this board?
Does it use the HIP6301 or what?
I am able to bench with 255x10. But I wish I could use these setting 24/7 :(
IMO I need more vdimm... mod anywhere?
Then I am gonna cool my tinies (mosfets) down so that I am able to use more than 1,85VCore w/out reboots...
My cpu could do 242x11,5 stably and I am convinced it still can do it with enough voltage... :)
I put a New UN-locked 2500+ in the Infinity last night and put active cooling on the NB & SB.
Its at home Priming at 245x9.5 to get it warmed up for some testing tonight.
nice.. keep us informed!
HI
HOw about the 10 - 10.5 multipliers ?
They are functional, right ?
Sure... why wouldnt they be??
Hi
Because they arent for those of us with NF7-S's.
I may order one.
Would a Northpole or Swiftech cooler be advisiable for the nb ?
I have Buffalo pc3700 BH5's...and want to really push the fsb.
I put a Vantec Iceburg on the NB & the SB.
Hi
Thats what I have on this Abit. I can just move it over.
The Infinity should be here on Tuesday
So far it feels alittle more stable... but I have not really tested all the new changes yet.
why is it limited to 1.95v
is it a bios issue or vcore regulation/overvolt protection?
??? For some reason when you try and use 2volts it becomes less stable?
That's it. But I can use it now afterupping my modded 12V line more and putting some sinks to the mosfets :)Quote:
Originally posted by clockedOut
??? For some reason when you try and use 2volts it becomes less stable?
Can go into windows at over 2,8GHz with my barton and bench up to 2780MHz :D
But wanna have that 100% stable ;) Vmod in progress, but I am afraid... I just have to hook it up and turn the sys on... but if it looks like :explode: I will :brick: !
NFII Infinity: ordered from GameVe :D
Hi
IS there a vmod for the weakish vcore ?
Does this board have the 4 mounting holes for mounting waterblocks or more importantly phasechange ;) ?
some chips just dont like voltages above 1.90v.. i have a
tbred 2100 like that now..
my 1700 wont go at over 1.85-1.875
1.825 is kinda iffy but will work. its a very picky chip. im hoping my barton will not be.
I asked podunk from over at broadbandreports.com if I could post his NFII Infinity bench. Yet another great 'clock on this board with a plus (IMO):
Done with Samsung TCB3 PC2700@3.0v - really nice stick there, podunk!
the above was done with....
dfi infinity
1800 dlt3c
1 stick 512 pc2700(samsung tcb3)
vdimm 3.0
vcore 1.85
vdd 1.8
this is the tighest i could get the timings. it would boot into windows at 230x10 @ 11-3-3-2.5. any more fsb and it would bsod.
does anyone know what would be the max vdimm voltage to run through this stick for benchmarking? 3.0 vdimm was just warm to the touch.
i'm just curious to see if there is any headroom left.
Hi
Is there a mod yet, for boosting the vcore ?
hi guys im stuck around 225ish dc
my hyperx just doesnt wanna do it...ive given it 2.9-3.3v...i even have the heatspreaders off (week 316 bh-5) with a fan blowing on them. I tried stick 1 and maxed at 225...stick 2...i dunno what it can go to, although it does hit 229. ive also tried many timings...agp voltages...vdd and such. If i put some 1x512 value ram in there its stable...ill have to see how far it will go with the value ram cause im sick and i havent had time to test much as the mobo was recieved today. it has to be ram cause ive tested different ram and its stable....now how to get this hyperx higher? im lettiing it burn in at 220fsb 3.2vdimm overnight as it is only 1mos old
btw i have 2 of these mobos so i can try another one, but i highly doubt its the mobo as its hit 229 stable. (so far ;))
Quick question :
The DFI has an option for Vdimm of up to 3.3V right ? If we were to set it at 3.3V, does it fluctuate or go as low as 3.1V during load ? I understand we need to mod our 3.3v rails in order to get a good stable 3.2Vdimm at load
correct. its best to have .2 v on your 3.3 over what your vdimm is.
so if we set it at 3.3vdimm then we need 3.5 on our 3.3 rail.
UNLESS DFI got smart and made the vdimm use the 5v line instead. that would be a very good move on the part of an OCer aimed board. this would allow much higher and much more stable vdimm and such.
Thanks a bunch Evil_Spork :D :D :D
Wait a tick..
ClockedOut was that 3dMark run @ 10x250?
Because @ 11x223 I'm killing those system scores... ? Perhaps you just have an old OS. Still quite a stellar bench! 20K is always a nice number to achieve. :)
SAE:
Is the DFI Infinity II slower per clock @ 3D compared to the NF7-S?
Mine is going under the soldering iron tomorrow night and I am expecting high 240's - 250 FSB when through. By the looks of your board though with the right memory this thing could touch the 270's with CPU-I enabled and 2x channel.
I simply think the NF7S has plenty of bugs
yea dominic my 3.3v rail is slumping to 3.2...i dont know why,,,its a true power, and on every other mobo, its 3.3v on the dot. perhaps thats why i cant get my damn hyperx stable above 220fsb
Hallowed, Maybe it has something to do with the difference between your 3200 & my 1700?? Or the fact that my 9800np doesnt clock to well.
Quote:
Originally posted by Hallowed
Wait a tick..
ClockedOut was that 3dMark run @ 10x250?
Because @ 11x223 I'm killing those system scores... ? Perhaps you just have an old OS. Still quite a stellar bench! 20K is always a nice number to achieve. :)
SAE:
Is the DFI Infinity II slower per clock @ 3D compared to the NF7-S?
Mine is going under the soldering iron tomorrow night and I am expecting high 240's - 250 FSB when through. By the looks of your board though with the right memory this thing could touch the 270's with CPU-I enabled and 2x channel.
any1 know of a utility besides mbm5 that shows your vdimm and vdd? mbm5 doesnt work to great cause the dfi infinity isnt listed. i need to figure out if my 3.3v slumping to 3.18ish is making my vdimm go up and down...therefore explaining why my damn hyperx cant hit 220fsb with out passing superpi 8m test
Wasnt reading. *slap* Thought you had a Barton.
At any rate, nice score once more. :)
Don't think the utility would matter...they all read the same sensor. So check it with a multimeter.Quote:
Originally posted by Daishi
any1 know of a utility besides mbm5 that shows your vdimm and vdd? mbm5 doesnt work to great cause the dfi infinity isnt listed. i need to figure out if my 3.3v slumDonping to 3.18ish is making my vdimm go up and down...therefore explaining why my damn hyperx cant hit 220fsb with out passing superpi 8m test
mbm5 doesnt work...id rather use a utility....any1 know of one?
sure is,Quote:
Originally posted by lght1
Hi
Is there a mod yet, for boosting the vcore ?
HIP6301:
Pin 7 to pin 9(Gnd) with 10K variable resistor for upto 2.2VCORE I believe.
Hi
Is there any pictures?
Yeah, a pic of the mod would be cool ! :)
Hi
I just installed mine tonight. Its been priming at 10x241 for several hours.
Too bad the 9.5x250 wasnt primable as the bandwidth was fantastic.
This is with the 12/18 beta bios.
Its wonderful to have the 10-10.5 multipliers back once again.
Which is the best BIOS for now ? Where could I obtain them ?
pulled from the amdmb forums hereQuote:
Originally posted by Swiftech
Which is the best BIOS for now ? Where could I obtain them ?
Oh you bastards... man I gotta sell this AN7 pronto.
damnit... I dont wanna buy any more AXP mobo's... already have 2 x NF7-S 2.0's running Seti all day...
right now I cant figure out if its my KHX3500C2 that wont go past 222fsb @ 2-2-2-6 timings with 3.0v Vdimm or the board that wont go any further... although I AM only using 1.7vdd... think thats the prob?
KHX 3500 is the exact memory that had problems with the NF7-S 2.0
What revision of the infinity have you guy's got ? and the north-bridge ??
Thanks Nomjr :D
There are a couple of BIOSes there so which is the best one ?
Btw, the NF7S is no match for the DFI
Hi
So far, this board is just doing everything so well. Just the simple fact of having those Abit denied multipliers allows for this Barton to finally shine.
I am really pleased with my purchase of the Infinity.
What kind of fsb speed are you guy's running ? i'm running 240 24/7 on my abit but i would like a 255+ 24/7so i think i'm ginna try the DFI Ininity.Quote:
Originally posted by lght1
Hi
So far, this board is just doing everything so well. Just the simple fact of having those Abit denied multipliers allows for this Barton to finally shine.
I am really pleased with my purchase of the Infinity.
:-D This board is total crap without the 12-10 bios...it hates bh-5 without the bios. Flash using winflash...its so easy.
Hi
I am currently using the 12/18 bios. Do you feel my BH5's would do better with the 12/10 ?
I do not know as i have not tried the 12-18. I think 12-10 is the favor among a lot of ppl.Quote:
Originally posted by lght1
Hi
I am currently using the 12/18 bios. Do you feel my BH5's would do better with the 12/10 ?
I am using 3.3vdimm to get 250fsb. I wouldnt run 3.3vdimm 24/7...but for benching its fine...just make sure the heatspreaders are OFF becaues they make no contact and have a fan blowing on them. If it craps out...it shouldnt...then you can rma it ;).Quote:
Originally posted by podunk
the above was done with....
dfi infinity
1800 dlt3c
1 stick 512 pc2700(samsung tcb3)
vdimm 3.0
vcore 1.85
vdd 1.8
this is the tighest i could get the timings. it would boot into windows at 230x10 @ 11-3-3-2.5. any more fsb and it would bsod.
does anyone know what would be the max vdimm voltage to run through this stick for benchmarking? 3.0 vdimm was just warm to the touch.
i'm just curious to see if there is any headroom left.
For how long can you run prim95 ?Quote:
Originally posted by Daishi
I do not know as i have not tried the 12-18. I think 12-10 is the favor among a lot of ppl.
I am using 3.3vdimm to get 250fsb. I wouldnt run 3.3vdimm 24/7...but for benching its fine...just make sure the heatspreaders are OFF becaues they make no contact and have a fan blowing on them. If it craps out...it shouldnt...then you can rma it ;).
havent tried yet ;). It can play cs for hours and loop 3dmark...im pretty sure its stable prime95Quote:
Originally posted by Astennu
For how long can you run prim95 ?
I could play CS verry long 3dmark no prob bud C&C Generals within 1 hour a hardware error. 240 is my max.Quote:
Originally posted by Daishi
havent tried yet ;). It can play cs for hours and loop 3dmark...im pretty sure its stable prime95
Here's another 250. Using XMS3500 and 3.1Vdimm, 1.8 Vdd. Haven't run prime yet but it's been folding for last 4 hours. Not pushing the CPU until I find out the max :)
Nice i have also got XMS 3500. What kind of NB cooling ?Quote:
Originally posted by dod
Here's another 250. Using XMS3500 and 3.1Vdimm, 1.8 Vdd. Haven't run prime yet but it's been folding for last 4 hours. Not pushing the CPU until I find out the max :)
The stock Zalman looking thing. Got a 120mm blowing in the side but that's it:DQuote:
Originally posted by Astennu
Nice i have also got XMS 3500. What kind of NB cooling ?
I think I need some advice from you guys though. Bandwidths not as high as I would have expected. Running Cas2/3/3/11
Edit: Ah, it's not 3dm2001 stable :( Crashed on second loop.
This cant be right i think your cpu interface is disabled you should have arround the 3800 mb/s @ 250 mhz. Try running 1.7 VDD and higer mem volts i heard that the nf2 chip could bugs out @ 1.8 ang higher maybe you have the same prob.Quote:
Originally posted by dod
The stock Zalman looking thing. Got a 120mm blowing in the side but that's it:D
I think I need some advice from you guys though. Bandwidths not as high as I would have expected. Running Cas2/3/3/11
Edit: Ah, it's not 3dm2001 stable :( Crashed on second loop.
Good spot. Interface was at optimal.
Interesting though . Using 250 FSB and 8.5 multi bandwidth is about 3580.
Dropping multi to 8 and 1.8 Vdd 3.2 Vdimm gives an increase as shown. Dropping Vdd to 1.7 gives no increase, result identical.
Enabling or Setting CPU Interface to it's most Aggressive setting improves memory performane.
Indeed and lowering volt will not give more mb/s it might give uyou more stability.Quote:
Originally posted by Drisler
Enabling or Setting CPU Interface to it's most Aggressive setting improves memory performane.
This is the 12-18 beta bios fixes...
BETA BIOS: N24IDC18.bin
Explanation for the file name of the BIOS:
"N"Force "2" Ultra "4"00 "I"nfinity "D"FI
"C" (12 in hex, which means the twelfth month of the year) "18" (Date 18)
Disclaimer :
This is a BETA BIOS for the DFI NForce2 Ultra Infinity motherboard.
Please use it at your own risk. DFI is not responsible for any damage
as a result of using this beta BIOS.
Key Changes Since The 11-27 BIOS (11-27 BIOS is an OFFICIAL BIOS)
1. Added the 12.5X multiplier to enable more aggressive overclocking
2. Improved DRAM compatibility when dual channel DRAM is populated
in the DIMM 1/3 slots instead of the DIMM 2/3 slots
3. Enhanced DRAM overclocking ability when the memory used requires
3.0V or above to perform at high FSB with satisfactory stability
4. Updated the Silicon Image 3114 BIOS to the latest 5.0.31
5. Changed the Optimal and Aggressive CPU Interface Table above 200 MHz FSB:
- OPTIMAL CPU Interface has been modified to be more user-friendly
for high FSB overclocking
- AGGRESSIVE CPU Interface now recalls the lost bandwidth when FSB
is above 200 MHz and at a multiplier that is under 10
Looks nice :D DFI is realy doing a good job for the overclockers :toast:Quote:
Originally posted by clockedOut
This is the 12-18 beta bios fixes...
BETA BIOS: N24IDC18.bin
Explanation for the file name of the BIOS:
"N"Force "2" Ultra "4"00 "I"nfinity "D"FI
"C" (12 in hex, which means the twelfth month of the year) "18" (Date 18)
Disclaimer :
This is a BETA BIOS for the DFI NForce2 Ultra Infinity motherboard.
Please use it at your own risk. DFI is not responsible for any damage
as a result of using this beta BIOS.
Key Changes Since The 11-27 BIOS (11-27 BIOS is an OFFICIAL BIOS)
1. Added the 12.5X multiplier to enable more aggressive overclocking
2. Improved DRAM compatibility when dual channel DRAM is populated
in the DIMM 1/3 slots instead of the DIMM 2/3 slots
3. Enhanced DRAM overclocking ability when the memory used requires
3.0V or above to perform at high FSB with satisfactory stability
4. Updated the Silicon Image 3114 BIOS to the latest 5.0.31
5. Changed the Optimal and Aggressive CPU Interface Table above 200 MHz FSB:
- OPTIMAL CPU Interface has been modified to be more user-friendly
for high FSB overclocking
- AGGRESSIVE CPU Interface now recalls the lost bandwidth when FSB
is above 200 MHz and at a multiplier that is under 10
that bios blowed for me...i couldnt even post at 2-2-2-11 timings. bleh i hear its good for ocz ram. Any other bios's out there i could try?Quote:
Originally posted by clockedOut
This is the 12-18 beta bios fixes...
BETA BIOS: N24IDC18.bin
Explanation for the file name of the BIOS:
"N"Force "2" Ultra "4"00 "I"nfinity "D"FI
"C" (12 in hex, which means the twelfth month of the year) "18" (Date 18)
Disclaimer :
This is a BETA BIOS for the DFI NForce2 Ultra Infinity motherboard.
Please use it at your own risk. DFI is not responsible for any damage
as a result of using this beta BIOS.
Key Changes Since The 11-27 BIOS (11-27 BIOS is an OFFICIAL BIOS)
1. Added the 12.5X multiplier to enable more aggressive overclocking
2. Improved DRAM compatibility when dual channel DRAM is populated
in the DIMM 1/3 slots instead of the DIMM 2/3 slots
3. Enhanced DRAM overclocking ability when the memory used requires
3.0V or above to perform at high FSB with satisfactory stability
4. Updated the Silicon Image 3114 BIOS to the latest 5.0.31
5. Changed the Optimal and Aggressive CPU Interface Table above 200 MHz FSB:
- OPTIMAL CPU Interface has been modified to be more user-friendly
for high FSB overclocking
- AGGRESSIVE CPU Interface now recalls the lost bandwidth when FSB
is above 200 MHz and at a multiplier that is under 10
check here...http://www.midnightmachine.com/modul...wcat.php?cid=2
member over at amdmb put this together
Running TwinMOS/Winbond BH-5 here and that 12-18 BIOS is fantastic. Sorry about your troubles DiashiQuote:
Originally posted by Daishi
that bios blowed for me...i couldnt even post at 2-2-2-11 timings. bleh i hear its good for ocz ram. Any other bios's out there i could try?
Thanks for the link, Podunk!
iirc that 11-19beta bios is not listed on the above link..i'm not sure what the 11-27 like. i havn't tried it.
hmm...maybe ill give it a try...Quote:
Originally posted by podunk
iirc that 11-19beta bios is not listed on the above link..i'm not sure what the 11-27 like. i havn't tried it.
What is the max you can run with that bios ? I wanna know if i would be able to get a 100% stable 250 fsb with my Corsair XMS PC3500.Quote:
Originally posted by sjohnson
Running TwinMOS/Winbond BH-5 here and that 12-18 BIOS is fantastic. Sorry about your troubles Diashi
Thanks for the link, Podunk!
i was using the 12-10 bios and at 250 it corrupted my hdd.. cold the bios be the reason? should i have used the 12-18?
Astennu
The 12-18 Infinity BIOS has settings to 300 FSB. I'm stable at 245 currently and can bansai to 250. This is with TwinMOS PC3200/BH5 and a rig that's running an old Win2k O/S load. I built it less than 24 hours ago.
Possibly memory issues aside, 250FSB stable seems a reasonable goal.
I think it has more to do with a to high oc but a other bios could help to solve the problemQuote:
Originally posted by Evil_Spork
i was using the 12-10 bios and at 250 it corrupted my hdd.. cold the bios be the reason? should i have used the 12-18?
I have seen people run 266 1:1. My current mobo ( NF7-S ) Can do 245 but it is not stable enough. The best now = 3 hours+ @ Prime 95 @ 240 FSB. it's realty the mobo of mem but i think my mobo. What kind of mem voltages are you using ?Quote:
Originally posted by sjohnson
Astennu
The 12-18 Infinity BIOS has settings to 300 FSB. I'm stable at 245 currently and can bansai to 250. This is with TwinMOS PC3200/BH5 and a rig that's running an old Win2k O/S load. I built it less than 24 hours ago.
Possibly memory issues aside, 250FSB stable seems a reasonable goal.
At 3.2 now but I want to lower the volts. Stable for 3dmark runs and prime95 stable for an hour. This memory has never really been burned in, so I'm jazzing it up just to see what this board has in it.
Now that I know I'm starting over. First, decent heatsinks on every hot and warm chip. Then, some PSU fine-tuning on the rails. Followed by some voltage-stabilization and then I'll begin burnin on the bh5...
Ordered mine today. If they ship it out tommorow it will be here monday and Ill see what she can do. Hopefully Ill be able to 3dmark at 250fsb but I will be happy with 240. Havent had this Corsair XMS 3500 BH-5 over 220 fsb as of now. Not sure whats limiting me the mobo or them memory as my 8rda+ only goes to 2.9 (2.86 actual) and the vdimm mod is hella hard. Hopefully my mem can get that high with some volts. How high should I crank up my 3.3v rail if I plan to be running 3.2v (or maybe even 3.3 if thats not enough) Im going to figure out what each pot in my fortron 530w does this weekend and crank them up. Right now my 5v is at 5.05 and 12v is at 12.47. Not sure what the 3.3v is because I havent jumpered the PSU yet to test.
How high you set your 3.3 depends on how low it dips under load. You don't want it to go below 3.3 under load, some would say not under 3.4 under load to compensate for unexpected demands.Quote:
Originally posted by Aldus
...
How high should I crank up my 3.3v rail if I plan to be running 3.2v (or maybe even 3.3 if thats not enough) Im going to figure out what each pot in my fortron 530w does this weekend and crank them up. Right now my 5v is at 5.05 and 12v is at 12.47. Not sure what the 3.3v is because I havent jumpered the PSU yet to test.
3.3v rail doesnt matter on the dfi...
vdimm runs off the 5v rail. Im almost 100% sure...my 3.3v rail is only 3.18 (issue with board), and my vdimm is perfectly stable at 3.3vdimm. Theres no way that you can have 3.3vdimm stable off a 3.18v 3.3v line...it MUST be off the 5v line.
i concur, i dont know whatthe deal is, but no way its on the 5v rail
Hrm I jumpered the PSU and took the readings again and got some weird results on the 12v line. While the psu was running by itself I got a reading of 11.00 on the 12v line. I was like wtf so I just left it be. But then after putting the comp back together and starting it up it reads 12.48. Very weird.
The 5v was the same as while running, 5.05. Im going to play with the pots this weekend and get it up to around 5.2.
The 3.3v was at 3.43. This is a nice psu. Havent had any problems with it or adjusted any of the pots yet.
Hopefully the board will be here monday so I'll have a week off of school to play with it.
Switching PSU's can often show different voltages on no-load versus loaded conditions.
It's actually hard on the PSU to jumper it to run without load. You can do it, but keep the time to a minimum.
The fact that the DFI can run good Vdimm on apparently low 3.3 rails could be due to it using the 5v or 12v rails to generate the Vdimm, or it could just be that the board doesn't report the 3.3 correctly. Measuring with a DMM at the ATX connector and on the motherboard (where? I don't yet know) is the only good way to know what the 3.3 rail voltage actually is.