Unlocked Venice 3000+

Printable View

Show 100 post(s) from this thread on one page

10-04-2005, 06:26 PM
Ugly n Grey

Quote:

Originally Posted by Daved+

IMO this isn't about money, it's a challenge, it's the joy to make something that seems to be impossible happen.

By all means, whack away, I'm not belittling the effort, just doing my part to point out the alternative viewpoint and solution.

I HAVE to admit to the ultimate puzzlement I have about all this though... what possible difference could it make? Our boards handle higher HTT speeds than the chips multis can reasonably produce, there is no such thing as a 1:1 memory setting on this platform. So what exactly is the gain if by some chance there is success?

As well, the M2's are coming out next year and that is creeping up on us fast, this will be a dying platform in no time. It's like buying an AXP off Ebay and modding it show up as an MP... fun but a waste of time considering some dude with an El cheapo x2 is putting up four times the numbers....

My honest advice? buy one of the Opty's that everyone and their dog seems to be cranking 3GHz out of and spend your spare cycles beating up the guys who bought Sandys.....

Anyhow, as I said, the geek in me finds this fascinating, please continue.

UnG
10-04-2005, 07:18 PM
Daved+

Quote:

Originally Posted by Ugly n Grey

BI HAVE to admit to the ultimate puzzlement I have about all this though... what possible difference could it make? Our boards handle higher HTT speeds than the chips multis can reasonably produce, there is no such thing as a 1:1 memory setting on this platform. So what exactly is the gain if by some chance there is success?

UnG, last time I checked, not every mobo can do sick high HTT and not every memory divider works for everyone. So having higher multipliers may be useful for some people ;)
10-04-2005, 07:38 PM
Postal Dude

David+: thank you dude. My Neo2 Tops out at about 270 on the HTT. SO no matter what chip i get, i cant get over 2.7Ghz. Experience with this winnie says that anyway. Not tried another chip. A higher multi would be great. scared to get a 3000+ or a 144 cus of the x9!
10-04-2005, 07:45 PM
Ugly n Grey

Fair enough Daved and Postal, I was thinking the DFI boards and you set me right. I can see a use for it there.
10-04-2005, 08:01 PM
jahjahbinks

So after 31 pages, no one suspected software bug?
10-05-2005, 02:12 AM
gbomb944

I'm surprised nobody unlocked one by now. I don't think there are even any unlock mods for 754 so apparently amd is good at locking them down.

The thing that somebody posted by the linux guy was interesting. He told the other guy that if the cpu "oopses" not to publish it so he seemed to think it was possible that what they were doing could cause an unlock. At least thats what I thought he meant.
10-05-2005, 02:57 AM
TiN_

I still have no progress because all things you have posted i tried

1. Enabling CnQ and playing with HTT (up to 330)/multi(x4-x9)/Voltage(0.8V-1.8V)/Freq's (800Mhz-2800MHz) ... Nothing
2. Renaming CPU String from registers C000003x (nothing as it seems to me prev.)
3. Play with pins (isolating each NC_xx, JTAG, STRAP_xx)

And i didn't get any HW still. I'll wait some more. I'm still hoping to use a JTAG-port on CPU. If it is ever RO, it is possible to write SOME data to it, because CLEAN register is 111111.... and when it flashed ..it become 1000110...etc...so we can put 0 if there is 1 :) But with multi this means that we can get x5 max multi from x9 cpu :) So I will wait for mobo arrival...

THX. (And don't flame such things as renaming cpu in win :)))))

UnG, BTW I think that locking sequence on ALL A64 (from Sempron 3100+ to Opty 8xx) is the same. They all have JTAG. So If people would get Unlocked Opty this will be perfect :) or an Athlon 64 X2 FX :)
10-05-2005, 03:25 AM
4rory

which stock multi's would be useful if we could write zeros over one, have you checked each one out and see if it was even worth it ?

Maybe amd already thought, first 0 means locked, rest of numbers mean from what multi.
10-05-2005, 03:44 AM
5days

CNQ + Renaming CPUID ftw?
10-05-2005, 02:36 PM
ixtapalapaquetl

Someone at DFI-street says he has an unlocked s754 3000. Thought you guys might want to get ahold of him.

http://www.dfi-street.com/forum/showthread.php?t=25811
10-05-2005, 02:47 PM
HiJon89

Don't believe there is a S754 Venice so something doesn't sound right :confused:
10-05-2005, 03:01 PM
Jesus

Quote:

Originally Posted by Vapor

*warning fart*

I bet permanent marker would work...pencil would probably come off with the AS5. ;)

/crap

:lol2: :rofl: :rofl: :bounces:

Good one!

Quote:

Originally Posted by jahjahbinks

So after 31 pages, no one suspected software bug?

probably but, maybe it's possible to reproduce that bug.

Something must have caused this bug, to stay. Maybe even a voltage peak.
10-05-2005, 03:38 PM
i found nemo

Quote:

Originally Posted by HiJon89

Don't believe there is a S754 Venice so something doesn't sound right :confused:

new semprons
10-05-2005, 03:44 PM
HiJon89

Quote:

Originally Posted by i found nemo

new semprons

You mean this?
http://www.newegg.com/Product/Produc...82E16819104241
10-05-2005, 04:21 PM
Thorry

Dudes, understand this:

Socket 754 CPUs are exactly the same as socket 939 CPUs, the difference is the memory controller. Since the unlocking or multiplier setting voodoo has absolutely nothing to do with the memory controller I see no reason why the magic shouldn't work on S754 cpus.

Also I don't believe the unlocking has anything to do with it being a Venice core, could be just a little screw-up at the AMD manufacturing plant and it just happen to be at the Venice cores. If the same machine processed some other cores later the same problem could exist in those other cores.

I have to totally agree to UnG, the unlocks are screw-ups at AMD. Trying to unlock it via the JTAG connectors has really no chance of working seeing as those have been on AMD cpus for like ever.

Trying to unlock the cpus won't work, people have been trying on Intel cpus since they were locked with no succes.

What people should do is write down the batch codes, good chance the batch codes are the same or very close together. At AMD they are now all in a folder called 'screwups'
10-05-2005, 04:46 PM
Clocker

Quote:

Originally Posted by Thorry

What people should do is write down the batch codes, good chance the batch codes are the same or very close together. At AMD they are now all in a folder called 'screwups'

These were the markings on the unlocked chip I had:

ADA3200DAA4BW
LBBWE 0531 EPDW
2735016 H50628

CxP
10-05-2005, 05:02 PM
i found nemo

Quote:

Originally Posted by HiJon89

You mean this?
http://www.newegg.com/Product/Produc...82E16819104241

no those are palermo cores, winnie based ... venice semprons are the ones i think the e6's with x64 and sse3 i'm pretty sure but could be wrong
10-05-2005, 05:09 PM
nn_step

I believe you are right Nemo..
10-05-2005, 05:12 PM
i found nemo

nope i was wrong e3

edit here's link

http://www.xtremesystems.org/forums/...978#post886978
10-05-2005, 06:45 PM
KoHaN69

Quote:

Originally Posted by TiN_EOF

But with multi this means that we can get x5 max multi from x9 cpu :)

So... you unlocked a lower multi? :confused:
10-05-2005, 08:18 PM
largon

Quote:

Originally Posted by ixtapalapaquetl

Someone at DFI-street says he has an unlocked s754 3000. Thought you guys might want to get ahold of him.

http://www.dfi-street.com/forum/showthread.php?t=25811

His proc isn't unlocked. Unlike nF4, nF3-bios' shows all multies, BUT higher value than defaults won't work.
10-06-2005, 01:46 AM
gbomb944

Has anybody seen the unlocked venice that started all this? I understand it's sold but to who? Whats the odds it was a hoax?
10-06-2005, 02:24 AM
Clocker

:slapass:
10-06-2005, 02:58 AM
Mehran

What?!?!?!?!?!!?!? What did u do with that?!
10-06-2005, 04:32 AM
Ref

Quote:

Originally Posted by gbomb944

Has anybody seen the unlocked venice that started all this? I understand it's sold but to who? Whats the odds it was a hoax?

It hasn't been sold - it is still for sale in For Sale forum.

Show 100 post(s) from this thread on one page