Printable View
Hello
Great thread and thanks Rabbi_NZ for making it :toast:
Currently testing my new system since friday and I just went to 240mhz and now 250mhz with 3.3 volts using info from this thread, its stable in memtest #5 so far but had some trouble with "memory read" errors in CS:S but i think i fixed that with a command line :cool:
http://www.xtremesystems.org/forums/...id=25007&stc=1
This Twinmos UTT "50D" chips is some freaky stuff.
Did volt scale testing, one stick, 8,2,2,2.0 on my NF7-S vmod'd dimm/dd/tt. Clearly the NForce2 hits a wall and it is a chipset issue.
3.2-3.35 yeilds 245Mhz 1T 8,2,2,2.0,
3.15V yeilds 240Mhz,
3.04V yeilds 225Mhz,
2.93V yeilds DEAD @ 2,2,2 so 8,3,3,2.5 @ 205Mhz max.
A sharp voltage curve that comes to an abrupt end when it hits the NForce2 wall, A64 reports show another 10-20Mhz in the ram that the AthlonXP cant get at.
Indicative of CH5ness in that low voltage and 5,2,2,2 timing not doable. Counter-indicative of CH5 is that 8,2,3,2.0 does NOT add any Mhz on NForce2. Making me think the improvements they made took away some of the 2,3,2.0 character of CH5.
Also, this ram is SERIOUSLY heat sensitive. If I turn the side fan off, and let the chips warm up they error-out 20Mhz lower per voltage setting.
The NF7-S wont do 250Mhz in memtest without erroring, while the DFI Infinity will. So the wall comes down a little sooner on the ABit than the DFI.
Nice thing is the 245Mhz is the fastest results I ever got on a NF7-S in single stick mode (thanks to 1T). Im using TicTac's D26MantaXT with CPC ON for testing. CPC OFF didnt do a thing for max FSB. Ill have to switch in the 2nd stick for dual channel test next.
i hit wall at 248 and 245 with 2nd stick....
Just updated first thread... cheers to everyone that contributed :toast:
esdee, you raised the average for EVERY voltage level you tested at! hehehe
uwackme, if you could post a nF2 Tweaker screeny (jus the Tweaker, no desktop) it would be great. I will use it on the first page as a reference.
Quote:
Originally Posted by Rabbi_NZ
and i will raise it more ... already gained 1 MHz in every voltage level (up to 3.5 , now 271Mhz is benchable)
:toast:
where are you memtest stable with 3.5v?Quote:
Originally Posted by esdee
still 270 under 3.5VQuote:
Originally Posted by Rabbi_NZ
I mail you a snap Rabbi.
3 more sets in today, all hit same max... 245Mhz DFI Prime/3D stable, 240Mhz NF7-S Prime/3D stable... 2x 512M 8,2,2,2,2.0 @ 3.3V. Higher volts gets either no change, or errors (from heat no doubt).
I have every confidence in A64 these sticks will do 250-260-270, it is NF2 that has a wall not the ram.
uwackme get me 2 sticks please :D
First post is undergoing more changes... putting up screen shots of nF2 Tweaker, A64 Tweaker and 865 Tweaker so members can use them as reference.
If you want to send $$ paypal, I'd be happy to grab you a pr Tedy.
here is the A64 screeny you asked for RAB* :toast:
the second one is my A64..:) TREF=1560
also i beat my superpi score with the new DFI UT bios in the first screen :D
uwackem you bought those at newegg for 140$, right ? and all 4 modules were UTT?
Yep, Newegg's the only place ive seen them.
rabbi_nz my winnie is a 3500+ :D
Fixed mate :DQuote:
Originally Posted by esdee
There are now tweaker screenies from uwackme, aCidbAbY, esdee and myself. :toast:
I've set my VX timings exactly like Tony (bigtoe) did in this post:
http://www.bleedinedge.com/forum/sho...39&postcount=2
Got a whopping 80 Mb/s increase at 261 htt in Memtest (it was 3107, went up to 3187) with them.
Quote:
Originally Posted by mrlobber
3187 :eek: is that with your current setup in your sig?
HSP on ...
3.6V @ 272 (maybe i can pull a 273, ill find out tommorow!) DC ofcourse!
Rabbi_NZ sorry :D i make you update the first post every day :)
http://www.xtremesystems.org/forums/...chmentid=25126
Yeah :D Too bad it's only "memtestable", Prime and Superpi stable only at 233 htt (seems to be a dodgy cpu, a CBBID 0451 DPMW).Quote:
Originally Posted by aCidbAbY
Cant run 20 tests stable... Not even at std. speeds.... - Not even with 3,2V - instead of stock 2,8v
With DFI nF4 Ultra-D
I think its a due of to warm mosfets ;) - Well... I dont mind, not an issue while playing - Played 3 days at lan :D
Esdee, is that on DFI nF4? My one cranks out errors as soon as I up Vdiim higher than 3.4v.
mrlobber, cheers for that... I'll check it out when I have time for sure
esdee, it's because of you I get no sleep! heheheheh
I'll update your results tonight :)
Sorry rabbi_NZ :DQuote:
Originally Posted by Rabbi_NZ
@mrlobber
Yes that on the DFI nF4 Ultra-D Bios 218, i just put some HSP on so i made it to get them stable while benching
http://www.xtremesystems.org/forums/...id=25130&stc=1
should be the lowest unpatched 29sec :D
Btw no problem here benching at 3.4V+
Try to connect that molex and that floppy cable on the board, if you haven't already
I have that motherboard and 2 stick of VX...Quote:
Originally Posted by esdee
I don't plugged the floppy connector because on the manual the +5 and the +12 are inverted from the floppy connector...
i plug that all tha same???