Is there any way to determine that the wall is on the CPU and not the motherboard without having to test 6 different motherboards? :)Quote:
Originally Posted by Neon Biker
Printable View
Is there any way to determine that the wall is on the CPU and not the motherboard without having to test 6 different motherboards? :)Quote:
Originally Posted by Neon Biker
Hmm FSB wall Hack , MORE VOLTS,
Could it be the PCIE frequency needs to be raised or dropped also
Even if I maxed the voltage up to 1.6v in the BIOS, my vcore would only read actual voltage at 1.54-1.55 due to vdroop.Quote:
Originally Posted by mukmaster
In addition, I'm only able to increase NB voltage to 1.4v.
I think I need a different MoBo to test the CPU.
the lowest i have seen is my old e6600, less then 400 of fsb wall :(Quote:
Originally Posted by flytek
What chipset Rev are you using ?
it may be related to that my old p5b-d that was a c1 was stuck at 440-490 the c2 has no such issue.
Info that I could gather form this thread
http://img525.imageshack.us/img525/8310/untitledxb8.png
Actually....I think it could be the board. I HOPE it's not the CPU. ;)
also: E6420 / L651 B325 and I've never ran the chip at 1.6v vCore.
The max BIOS is 1.6, the max actual is probably 1.5-1.56v.
The max real voltage I've ever had on the chip actual is 1.535v.
.
.
I've tried 2 mobos, max FSB i got is 523, at bouth...that's strange.
Is there any pos to hack FSB wall?
Quote:
Originally Posted by mukmaster
well, I could add like 10 more cpus with walls at 500, 520, 545, 555, 470, 460.
All of them can be broken by xtreme cooling
MB chipset fsb wall can be broken by using memset MCH bar
This is going to sound stupid, but does the memset program work with the nVidia chipset?Quote:
Originally Posted by kiwi
I thought it was Intel only?
@mukmaster
380 is the max stable, but there is no way to post at 400 or over with any setting
my current Xeon 3060 has the fsb wall over 500 (never tried more), with the same P5B-D
E6300 ~470
E6600 <400
X3060 >500
Memset 3.3 beta2 Performance level ?Quote:
Originally Posted by kiwi
Quote:
Originally Posted by yotomeczek
No, look at pic :) But Felix removed it as far as I know
http://img248.imageshack.us/img248/4743/aaagd6.png
...if you want statistic: E6400 L630A407. FSB Wall around 565MHz with 6 multi @ P5D DLx C1(cooled by chiller +13 under load) and Commando(cooled by vapo ls). I tried different voltages but higher vcore, vfsb, vnb => more stable. With 8 multi - only 556FSB validation under vapo ls (dont know was it limited by cooling or fsbwall).
I'd love to get my hands on that.....;)Quote:
Originally Posted by kiwi
Hey I will be out forthe weekend but when I get back I will update my chart. :cool:
I cant find any relation between anything that could make a wall it just seems to be the chip. your either lucky or your not.
To the statistics - vcore - 1,6v (real), do 440x8 (can't do 440x9 because CPU max is about 3760 MHz) - tested on 6 mobos (P5B Deluxe, P5N32E, Foxconn 680i (I don't remember name of this mobo, as foxconn gives to complicated and too long names :ROTF: ), Ga P965 DS4 1,0rev, MSI P35 Neo-F, Biostar Tforce P965)Quote:
Originally Posted by mukmaster
Another CPU - E6400 - FSB wall 413 MHz, Vcore 1,7v (real) *crappy one* - tested on 3 mobos (P5B Deluxe, Foxconn nf680i, P5N32E)
E6300 - FSB wall 480 MHz, 1,5v (real) - tested on 2 mobos (P5B Deluxe, MSI P35 Neo-F)
I know this post is quite old, but try getting it at 400, instead of incremental changes. Its the chipset thats got a too high oc, but there's a strap change at 400FSBQuote:
Originally Posted by Viper.Br
my E6400 have a wall at 500 on air with P5b deluxe , and 502 on water with P5k .
i try to raising vcore but nothing help
well it seems there is a real problem with almost all the 200fsb defaulted cpus. massive overclocking potential is being lost due to these fsb artifacts. i think we need to figure out every single voltage and timing set/code applicable to the cpu's and the chipsets...
maybe p35/x38 from some clever manufacturer will have the necessary tools to improve the overclocking of cpu's that unexpectadly fail to OC high enough on fsb
I heard there was a new option in the P5k series along the lines of 200/266/333 fsb default. I cant test cause I have the P35-DQ6 instead.
i won 15 mhz on the max fsb on my E4300 by doing the ihs removal ... max frequecy also got 9x15=135 mhz more than with the IHS ... i haven't been looking on the temps , but i was fsb walled at 377 mhz before and now , it's seems to be around 395 , but i have to test more to find it
Im almost positive this is the case with the evga 680i... Im still playing around with my chip but Im almost positive this is the case with this board.Quote:
Originally Posted by kiwi
Going to try a 6x multi though.... hadnt thought of that...
farthest I got on 7x was 500fsb
http://valid.x86-secret.com/show_oc?id=216981
That sounds good ... did anybody test this ?Quote:
Originally Posted by megatron
yes! I was waiting for someone to mention the E4300 finally. Its all I could afford and on my DS3 rev3.3 I hit the FSB wall of around 370-375 REGARDLESS of the multiplier I used. And, I have read literally a hundred or more threads regarding this combo and I have come across exactly TWO people that have been able to got a E4300 to do 400 or more on the DS3. Where the E6XXX wil do 500fsb no problem on a DS3. I basically gave up on my E4300 and put it back to stock for quiet low power tower, when I can afford to I will try something from the E6XXX line.
If anyone feels charitable and wants to trade me a E6300 for a E4300 just let me know. Seriously, anyone want to do some testing on a E4300 and has a 6300 lying around Ill trade. Mines like a week 43 or somewhere around there.