View Full Version : Weird multiplier problems w/ Barton
Ok Ive had my new Barton for about a day now. Im still testing it and torturing. However... I went to go set it to do a prime test at 2.4ghz. And no multipliers above 11 will work. I wanted to do a 12x200 test. Computer specs are in sig. Just a different CPU.
Bartons arent locked.. so... why wont above 11 work? Am I missing something?
c627627
11-04-03, 11:18 PM
Abit and Soltek nForce2s read all multipliers on all CPUs without any wire tricks 4 sure.
Your Abit not posting all multis can happen sometimes. What multis can you do and what happens if you lower the FSB, can you still not use a specific multi?
You may need more Vcore or Vdimm voltage to hit 2400 MHz. You'll also need a Power Supply that can handle this above average overclock.
Dont worry man I gots all the tools. This sucker is stable at 2366 no problem. But the thing is I cant get a fsb above 218 or so. I havent put better cooling on the NB yet. I know it should be completely unlocked. But no matter what I set it to be it 11.5, 12, 12.5 or above it keeps it at 11. With my other CPU I could use multis above 11. Also I use a Sparkle 400W which is very good. My vcore was at 1.85 at the time and my vdimm is the same as my sig. Im sure it isnt anything like my ps, memory, or voltages.
I guess ill try 11x218, though thats cutting it a little close to my boards max. But the issue isnt whether I can get to 2.4ghz. Its why I cant use a multi above 11. Its very odd thought maybe someone else had a similar problem that they solved.
c627627
11-05-03, 01:55 AM
Can you do a specific multi with lower FSB?
Which multi's can you not do with any low FSB?
Goldlocke
11-05-03, 02:00 AM
IR1, maybe AQYFA stepping? serial "Zxxxxxxxxx"? week 42?
Yeah mine is a AQYFA 342. Why?
Also it doesnt matter what my FSB is when I try, any multi over 11. It just uses 11. Ive tried many combinations.
Edit:
I did a search in google. The only site I get is this Austria site. Gold maybe you know something about this stepping? Anyway Im gonna look through it... you sound like you know something Gold.
Crap... from what I can gather from this Austrian site. Overclockers.at these new 2500+'s the AQYFA week 342's are locked and dont allow anything above 11...
Wow... wtf AMD? ROFL Goldlocke I ran into some posts by you on that forum. I see the picture you posted crap so what, AMD is starting to lock again!?!?
Goldlocke
11-05-03, 02:45 AM
yes, I'm a resident of www.overclockers.at
2 users here in Austria ran into the same multiplier probs as IR1. Exactly the same stepping, same week. look at pic!
2 possibilities:
1.) maybe that week42 is just "broken" or...
2.) AMD REALLY DID IMPLEMENT a MULTI LOCK !?!?!?
IR1, you understand "Arnie german"? Greetings to "Gaaalifooornia" ;-)
Cause you slightly misunderstood my posts. I stated that according to the pic, there is no difference in Lasercuts from my unlocked AQZEA's or AQXEA's to the AQYFA. As far as I could see.
edit: I made a head to head pic -> click here! (http://www.overclockers.at/attachment.php?attachmentid=40189)
Heh yes we now have an Austrian Governor, which is cool, as long as Gray Davis never has to be seen again.
Yeah I did misunderstand. It was a bit difficult to comprehend considering I dont know Austrian and Google is not a perfect translator.
Yes I see that there is no outward physical difference between the AQYFA and the earlier steppings. Which is REALLY weird and bothersome... man I hope someone finds a way to fix this if it is a lock.. or maybe its ab error? I doubt its an error considering they would use the same chips for higher clocked processors like the 2600+... I bet its a new multi lock. Maybe connecting to pins underneathe will fix it...
I guess I didnt check.. but I cant even go below 11 multi. If I do my computer wont boot at all and I have to reset the cmos.
c627627
11-05-03, 05:17 AM
OK so all your multis are unavailable to post. You only have default multi of 11 available to you, that's putting it a little bit more clear than it was in the beginning of the thread.
Well we know it's not the mobo since you have Abit, which means if Abit nForce2 can't read all multis on this chip, most probably all mobos cannot.
So try the wire tricks like people who don't have Abit or Soltek have to for all CPUs (except sometimes for CPUs whos default multi is 13 and up, like 2100+.)
http://forum.oc-forums.com/vb/showthread.php?s=&threadid=232303
Goldlocke
11-05-03, 05:26 AM
...and there is no microcode diff.
cpuid-z or wcpuid show exactly the same family, model, stepping code.
6-A-0
IR1, I also think u should try the pin mod. lets see...u can use conductive ink.
dropadrop
11-05-03, 11:46 AM
Well, I'm just setting up a computer for a friend of mine, and have the same problem. Don't remember the excact week, but it was definatly over 40. It's a 2600+ barton with brown packaging. This really sucks as he did'nt get especially good ram, I was thinking he could overclock with the multiplier! :D
Anything over 11.5 does'nt make a change, and anything under does'nt post. (what I tested anyway)
c627627
11-05-03, 03:59 PM
With which mobo, dropadrop?
I still can't believe that only two nForce2 mobos (Abit & Soltek) read all multis but it appears to be so.
Now there's news that even Abit can't read all multis on new Bartons? The question is, do wire tricks work with Abit too?
But the thing is.... you cant change the multipliers at all. Im stuck at all. I thought I had changed my Multi to something lower than 11... before. But I guess I hadnt. Anyway yeah im stuck at 11. This is extremely weird. Ill look into doing the pinmod. But I hope some website figures it out soon...
c627627
11-05-03, 04:52 PM
I have an Epox 8rda+. It's an nForce2 and yes I'm stuck at default multi on any chip too...unless it has a higher default multi (13 and up, like 2100+, 2200+, 2400+ and Barton 3000+) Then I have all multis available to post.
Red multipliers in the right column here:
http://www.c627627.com/AMD/AthlonXP/
Please report back if the wire trick works on Barton-Abit combo.
Wire trick is a no go. Same problem after wire trick. This isnt looking good.
this is really alarming. I sure hope AMD does not do an Intel and lock the multipliers so that we can't change them...
have there been reportings of other steppings with locked multipliers of the Barton?
Goldlocke: thanks so much for your help!
WELCOME TO THE FORUMS!
c627627
11-05-03, 09:18 PM
This is a big deal.
New Bartons are therefore factory locked? I hope we can get further confirmation on this.
There arent any lazer cuts on top but do the traces underneath actually complete the circuit? I cant tell from the picture of the "locked chip".
http://www.theforumisdown.com/uploadfiles/0103/GB02.gif
Can any of the owners provide a clear pic of the L1 bridges of their supposedly locked bartons?
I've just been pointed to someone with a similar problem, but with a AQZFA stepping, in the Asus motherboard section: http://forum.oc-forums.com/vb/showthread.php?s=&threadid=243926
Strange thing is, there have been positive threads about the AQZFA.
For example, pacino running 10x255: http://forum.oc-forums.com/vb/showthread.php?s=&threadid=239687
c627627
11-05-03, 10:40 PM
Originally posted by Sonny: Can any of the owners provide a clear pic of the L1 bridges of their supposedly locked bartons?
If dates of production matter and if we're talking a recent "change,"
here's a week 41 Thorton (same as Barton with L2 bridge cut.)
AXDC2000DUT3C AQZFA 0341SPMW
http://www.c627627.com//IMG/Overclockers/L1Thorton.gif
P.S. Isn't TIFOSI some Euro soccer gang or smtng...
Edit: Note that pacino's Barton on which he could change the multi's is full 2 and a half months older than week 42.
Edit Edit: It doesn't matter.
Originally posted by c627627
http://www.amdforums.com/showthread.php?threadid=269919
says there's no outward physical difference between the new "locked" Bartons and old "unlocked" ones, how did AMD lock them then?
[/quote]Can any of the owners provide a clear pic of the L1 bridges of their supposedly locked bartons?[quote]
If you download the picture and enlarge it to its normal size. Or just have it enlarge in your browser you can clearly see the bridges. Here wait... ill download it, crop it, and then post a better close up.
Edit:
ok here is a close up
http://ianr.hypermart.net/half/aqyfa.jpg
dropadrop
11-06-03, 02:40 AM
Originally posted by c627627
With which mobo, dropadrop?
I still can't believe that only two nForce2 mobos (Abit & Soltek) read all multis but it appears to be so.
Now there's news that even Abit can't read all multis on new Bartons? The question is, do wire tricks work with Abit too?
This is with an abit nf7-m. (newest bios)
Samildanach
11-06-03, 02:55 AM
I can confirm that the wire trick on the socket as described at www.ocinside.de will not unlock my Barton 2500+ AQZFA on my Asus A7N8X Deluxe Rev 1.04. :mad:
Have tried multiple bioses: from1.04-1.07, including the uber bios. No change.
Shiyan has kindly linked to my earlier post. Go there to see my specs etc.
Short of actually blowing the bridges and/or sivler painting them I'm fast running short of ideas. :(
Anything else you people can think of?
Samildanach
blazer99
11-06-03, 10:29 AM
Has anybody had this same prob with a Rev.2 A7N8X board? Its my understanding the rev.2 boards may be able to read 'em?
Goldlocke
11-06-03, 10:53 AM
Shiyan, thanx for the warm welcome. :)
I'm sorry that I couldn't provide a higher resolution pic of the AQYFA. But the 2 owners here in AT don't have good digicams. :(
Though both users stated that the L1 bridges on their AQYFA's look exactly the same like on my pic of the AQZEA.
Any positive multi - reports from AQYFA's yet? I don't think so.
btw. TIFOSI are italian Formula1 AND Soccer fans! ;)
blazer99
11-06-03, 01:31 PM
just completed a "test" on a 2500 barton AQZFA 0339 UPMW
Multipliers were locked
so i'm not buyin' that one!
Goldlocke
11-06-03, 02:52 PM
@ samildanach: Did you try THAT pinmod? http://www.athlon-unlock.co.uk/5202/5283.html
And it didn't work?
/\/\/\
Thats the one I tried and it was still locked after that.
Samildanach
11-06-03, 03:26 PM
To Goldlocke, I haven't tried the wire/pin mod exactly like that. What I did was connect the corresponding socket holes on the motherboard with standard bell wire and then inserted the barton. I would've thought that would do the same thing
Samildanach
Goldlocke
11-06-03, 03:42 PM
wow, pretty effective way of locking by AMD here! damn.
blazer99
11-06-03, 03:50 PM
Yippee...
I found an AQZEA 0332 and a AQXEA 0334
Think I'll take the AQZEA...its slightly older and from what I can find a super-clocker.
Any thoughts?
Jim
Check out this other thread. Apparently new Bartons have locked multis. :(
http://forum.oc-forums.com/vb/showthread.php?s=&threadid=244237
the unlocking method linked to (http://www.athlon-unlock.co.uk/5202/5283.html) unlocks multipliers up to 12.5x, so all the multipliers above that are still locked.
have you guys tried setting a certain multiplier using that wire method, rather than the unlock?
for example, try setting something like 12x? http://www.oc-inside.de/html/workshop/pinmod/amd_pinmod.html
if you set the multiplier to a certain value, does it then boot with that multiplier or does it still boot with 11x?
probably won't work, but I'm wondering whether it has been attempted or not.
frankiej
11-08-03, 02:03 AM
Ok I have ? heres list of my comp and what specs I know of
A7N8X Deluxe rev 2.0 c18d1007.bin
barton 2500+ AQUCA 0303XPAW 180x11 1.98ghz
xms3200 2x512 4-2-2-2 run at 100% for 180
PNY5900Ultra 525core 975 memory
wd1200jb
raidmax520watt power supply
liteon dvd/cdrom and cdrw drives
core voltage is on auto now but have tried to raise it and shows it fluctate as such according to changes in bios but does not help do anything about stability and only slightly raises temps, but will not even get to windows above 187 or 188 fsb in the first place so why even bother raising volatge when it runs stable as hell at 180 fsb without one crash....ah well if have any ideas or questions please feel free to share , good luck all
mongoled
11-08-03, 06:10 AM
Well I have the same problem here,
Here is a pic of the Barton I just bought
http://homepage.ntlworld.com/dresuk/Barton_AQXFA-0340MPM.jpg .
Im hoping the company i bought it from will replace it with a older stepping one, received the CPU yesterday morning sent them an email in the afternoon. Finger crossed.
Its a AQXFA 0340MPM
mongoled
c627627
11-08-03, 06:46 AM
AQXFA 0340 MPM was not on the locked list. It was added thanks to you.
http://forum.oc-forums.com/vb/showthread.php?s=&postid=2252582
I really don't know what the pattern is for locked CPUs.
I guess the only way to increase your chances of not getting a locked one is if it was made before week 39...
You sure this is multiplier locked to 11, right?
I'm putting you pic in the main thread if you don't mind.
mongoled
11-08-03, 07:10 AM
i just noted your comment abt having to save and exit twice. Im pretty sure i tried that without thinking abt it. I tried getting higher multipliers and lower multis with no change.
Glad my pic was usefull
mongoled
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.