conroe fsb wall hacking. please help

[fscussel]

I can partially confirm that FSB wall has been hacked A guy of this forum knows details but he won't share them yet

So fsb wall is possible to hack

a processor CPU wall you mean?? how?

[ownage]

Sounds like total BS to me. If a chip is at his limits, how the hell would you get highers FSB's?

[mrlobber]

I can partially confirm that FSB wall has been hacked A guy of this forum knows details but he won't share them yet

I can confirm the same As far as I know, though, it only applies to Asus P5B boards at the moment and involves a couple of hardware mods of the cpu power circuitry as well as modded BIOS although I'm not sure about that last one. The guy who supposedly has done it has showed us a couple of cpu-z shots of E6400 validations at 620+ fsb. According to information available for me he was using a 630A448 cpu - the same batch number as my E6400 which, as I've determined, has a fsb wall at around 490 Mhz fsb.

[n00ba]

mrlobber
good news.. maybe help me..
I have E6300 with gigabyte DS3, and fsbwall for my cpu is about 430 MHz. Other cpu on my system reach 495 mhz .. On my DS3 Vcca pin directly connected to Vpci-e , but change vpci from bios can't help me..
Vmod pci-e(and vcca) up to 1.8V also not to help to rise fsb.

[dinos22]

I can partially confirm that FSB wall has been hacked A guy of this forum knows details but he won't share them yet

So fsb wall is possible to hack

that fruitcake Tapakah just threw a towel in >>> he's lying

too bad i got my hopes up

[kiwi]

Fruitcake lol

OK but what is this then?
http://valid.x86-secret.com/show_oc?id=145996
http://valid.x86-secret.com/show_oc?id=143757

If Franc makes cpuz, maybe he can comment on this?

[s e t h]

well a lot has happened since this thread started (like gtl vmods and boards that can tune it) but i have a reason for dragging it up.

Kunaak has tested a pin/paint mod that defaults cpu's fsb to 333bus.
so far this has resulted in a 16mhz fsb improvment on his Q6400.

anyway what needs to be tested is if the fsb walls improve with this mod for conroe and allendale cpu's.

here's the mod. post #18 http://www.xtremesystems.org/forums/...d.php?t=132900

[FabricioGS]

i'm trying a 6600 with very low fsb wall near 380mhz, i have test this mb with a 6300 for 1 month RS at 470mhz and fsb wall near 490mhz.

Mine E6600 week 23 has a low 390 fsb wall and also needs a lot of vcore to be stable. I'm very frustrated

I'd be happy to reach 9*400 with stability and low voltage

[Pete]

The amount of v mods that are coming out is amazing. I'm lookign forward to getting my bord modded soon and having a small bord within the bord to have all the mods on just to show them off! Looking forward to this

[megatron]

the P5B deluxe has a NBCC that gets higher the lower you set the multiplier.

say you have your E6600 and are setting 6 x 500 cause you think that will work.

Multiplier divide by set multiplier = X, take X and multiply that by the FSB, then take that number and multiply that by 4 and you get the NBCC.

9/6 = 1.5 x 500 = 750 x 4 = 3000 mhz.

a 3 ghz NBCC is extremely high and in 90% of cases will not be bootable.

a CPU with a lower default multiplier has a far better chance at hitting 500 then a CPU with a high multiplier, which is why you see E6300's do it all the time, but never a E6700.

yes, the 4 meg cache will hold them back... but even a 4 meg cache CPU can hit 500 stable... so thats not the issue.
its the north bridge core clock.

a typical NBCC will not like going higher then about 2400 mhz before freaking out.

if your doing default multiplier, 2400 is enough for a 600 FSB.
but once you lower your FSB, you start to drastically decrease your CPU's ability to hit stable FSBs that are really high cause you greatly increase the NBCC faster.

This explaination fits the experiences I am having

I can run 375 x 8 but I cannot boot at 375 x 7! If u do the math the NBCC is running higher, thats the only difference. With the e2160 default multi is x 9.

Now I need a pin mod for the default multiplier. Doh they stopped that a while back!

[s e t h]

try with a 6x multi too and see if the theory still holds

edit: i think what you may need most is a pinmod for the cpu to set its default bus at 266fsb...

[megatron]

try with a 6x multi too and see if the theory still holds

edit: i think what you may need most is a pinmod for the cpu to set its default bus at 266fsb...

I did the pin mod but after booting at previous settings I set "load optimised defaults" and it would no longer boot. I removed the mod and I am running again. The doubt in your mind if u have killed ur cpu/screwed ur bios and the relief when it finally boots again is what makes O/cing fun, lol.

[hey]

so ?? hipro did u tried VCCA modding ? did it work or what ??

[s e t h]

I did the pin mod but after booting at previous settings I set "load optimised defaults" and it would no longer boot. I removed the mod and I am running again. The doubt in your mind if u have killed ur cpu/screwed ur bios and the relief when it finally boots again is what makes O/cing fun, lol.

lol...classic...soz for the stress...wierd how a 266fsb cpu could cause no boot situation though.
could it have been some sort of ram setting issue? - maybe too little volts for the ram booting at a possibly higher speed after the bios clear with the new cpu?
anyway one would think a 266fsb cpu would make no difference and its unlikly the cpu was being pushed too far for default voltage...although that may be the case too..

you have given me an idea for wires running from under the cpu and maybe a jumper for the mod though.

[saaya]

well a lot has happened since this thread started (like gtl vmods and boards that can tune it) but i have a reason for dragging it up.

Kunaak has tested a pin/paint mod that defaults cpu's fsb to 333bus.
so far this has resulted in a 16mhz fsb improvment on his Q6400.

anyway what needs to be tested is if the fsb walls improve with this mod for conroe and allendale cpu's.

here's the mod. post #18 http://www.xtremesystems.org/forums/...d.php?t=132900

thats a different fsb strap which, afaik, means the chipset uses more relaxed timings. thats why booting at a lower fsb and bumping it up in windows gives you better results than booting at a high fsb directly. this doesnt have anything to do with the mod discussed in here though.

[megatron]

lol...classic...soz for the stress...wierd how a 266fsb cpu could cause no boot situation though.
could it have been some sort of ram setting issue? - maybe too little volts for the ram booting at a possibly higher speed after the bios clear with the new cpu?
anyway one would think a 266fsb cpu would make no difference and its unlikly the cpu was being pushed too far for default voltage...although that may be the case too..

you have given me an idea for wires running from under the cpu and maybe a jumper for the mod though.

The cpu as it is is recognised by the bios as "patch 0b5" or something, maybe it just didnt recognise it anymore (p35-dq6). The other possibility is I fluffed the mod somehow.

[megatron]

try with a 6x multi too and see if the theory still holds

edit: i think what you may need most is a pinmod for the cpu to set its default bus at 266fsb...

I can boot at x6. I booted up to 395 fsb at x6 multi. So seems not to tally with the NBCC theory.

[Fr3ak]

The NBCC theory turned out to be wrong a good while ago.

But with some Intel CPUs running a lower than stock multi can cause very dodgy behavior, so there must be some kind of issue, thats for sure.

P35 based boards seem to be able to run a much higher FSB on CPUs than the same CPUs were able to on other chipsets.

[evoic]

So......if the highest I can boot my E6420 stable is around 8 x 445ish, then there's a chance that the CPU is to blame and not the Motherboard???

I checked using a different Multi and came up with the same problem.
Nothing above 7 x 445ish and I didn't test it with 6x.

Any clarification here would be awesome if the answer to my issue is obvious so someone more knowledgeable in the subject than myself.

Thanks!

[evoic]

Anyone?

[s e t h]

445 is insanely low for an fsb wall...the lowest i've seen personally is around 460fsb...but yes it is entirely possible.
usually actual cpu fsb wall is the last thing to suspects as it could esily be the board or ram holding you back.
only once those two have been eliminated as possibilities and the cpu has the same wall at all low multies can one start legitimately suspecting the cpu

[evoic]

I'm going to have to disagree with you in a huge way.
445Mhz being insanely low equates to 1780Mhz.....

Just in the last year or so have boards been spec'd at 1066Mhz / 1333Mhz.
For a long time, anything over 400Mhz FSB was considered to be the ultimate clocker, so I'm surprised to see you write that the lowest you've ever seen is 460Mhz???

Anyways, aside from that, I'm not exactly sure how to go about finding out if the wall is the CPU or the board itself....outside of putting another CPU in the system and seeing how high it clocks.
I know it's not the RAM.
I can run unlinked with this board, and the RAM itself has been up to 1282Mhz.

With the E6420, I have 3 options: 6x, 7x, or 8x multi.
445'ish is the limit, regardless of multi, so I can't determine from that whether the motherboard or the CPU is holding back the higher speed.

[Xtreme Addict]

445 is insanely low for an fsb wall...the lowest i've seen personally is around 460fsb...but yes it is entirely possible.
usually actual cpu fsb wall is the last thing to suspects as it could esily be the board or ram holding you back.
only once those two have been eliminated as possibilities and the cpu has the same wall at all low multies can one start legitimately suspecting the cpu

I have FSB Wall on 440 MHz... on 6 motherboards

[evoic]

I have FSB Wall on 440 MHz... on 6 motherboards

Then that would point towards the CPU being limited, correct?

[Neon Biker]

Correct.