AOA Forums AOA Forums AOA Forums Folding For Team 45 AOA Files Home Front Page Become an AOA Subscriber! UserCP Calendar Memberlist FAQ Search Forum Home


Go Back   AOA Forums > Hardware > AMD Motherboards & CPUs

AMD Motherboards & CPUs Questions or comments on AMD products?


Reply
 
LinkBack Thread Tools Rate Thread
  #1 (permalink)  
Old 25th March, 2003, 12:10 PM
Liquid3D's Avatar
Member
 
Join Date: December 2002
Location: Newport, RI SURF
Posts: 374

Has cause been determined for 333FSB CPU's inability to attain high FSB OC's?

[B]Every once in a while I come across someone who cannot overclock their 166FSB based processor above 190FSB speeds. Several months ago I came across an article in Sudhian entitled; Opening the nForce floodgates --- Modding the AthlonXP for a 400MHz FSB. I read the article, but still haven't found, or read anything which speaks on the specific cause of the anomally; only that it exists and there's a fix. The quote below will illuminate the scenario; "The advantage of the 2600+ / 333 model is that it runs a 12.5x multiplier, allowing any Socket A motherboard with the capability to adjust the board's running multiplier.  This makes the CPU particularly handy for testing motherboards, as it guarantees compatibility for testing, and allows for the multiplier to be dropped for higher FSB testing....We've been using the CPU for motherboard testing, but had noticed that none of the motherboards we were testing could push past 190 MHz.  The results, no matter the board, were the same.  Initial booting into Windows would be fine, as would basic programs, but any attempt to seriously benchmark the system would fail.  While that didn't seem abnormal at first, after numerous boards ran into the same barrier at the same speed, we began to suspect there might be another component holding us back.  After swapping out RAM, video cards, hard drives and every other component, we focused on the CPU itself and began searching for a reason why we might be having such problems.  Even purchasing another AthlonXP 2600+ / 333, however, didn't solve the problem, and we continued to run into problems around the 190-195 MHz range....This was infuriating.  The whole goal of being able to run at high FSB's was to test a batch of high-end memory we'd received.  While you can run the FSB and memory bus out of synch, it lowers overall performance and can invalidate stress testing of a part.  With all of our motherboards stuck at 190 MHz (or lower) it was impossible to tell just how much various PC3500 and 3200 parts truly differentiated from each other....We spent almost two weeks trying to find a solution until we ran into Randi (MrIcee) over at XtremeSystems.  Two of the forum readers there (Aceman and Praetereo by name) have found a solution to this problem, detailed in a forum thread: http://www.xtremesystems.org/forums/...;threadid=6934  (see page three of this post for pic of the L12 mod, as well) Randy passed it along to us, so thanks very much to the boys over at XS...The problem, apparently, is that the nForce2 (and possibly the KT400 as well) destabilizes when the front-side bus is overclocked using a 333 MHz FSB CPU.  The solution (which we've tested) is to cut one of the L12 bridges on the CPU" http://www.sudhian.com/showdocs.cfm?aid=307

It occurs to me the immaturity of the AMD 166FSB processor, hasn't revealed enough of these circumstances to warrant AMD technicians and motherboard chipset makers into a collaborative research project regarding the issue, and finding a proper fix. Perhaps AMD's attitude is, the small percentage of overclockers/enthusiasts whom push their processors to these "insane" FSB speeds doesn't justify the cost of such research? Or perhaps AMD feels it's mainboard chipset makers responsibility (being the anomally is in the FSB indicative of a mainboard chipset design) OR a juxtiposition of this attitude amoung mainbiard chipset makers, who feel the design flaw resides purely in the processor design. Either way as DDR400 is supported in KT400, and nForce2 chipsets I feel it's unfair to expect end-users to void their processor's warranties to exploit a "standard" feature mainboard makers clearly used as a selling/marketing point. 
__________________
(1.) 2.4C@3.6GHz 300FSB, SLK900U/Vantec Tornado 92mm/119CFM, Abit IS7-E, 1GB OCZ GolD Dual Channel 3700, Tyan G9700 Pro, PCPower&Cooling 400W
(2.) JIUHB1700, Prometeia, Soltek 75FRN2-L, 1GB OCZ EL-PC3700, Leadtek Winfast A250, ANTEC 480W


Last edited by Liquid3D; 11th April, 2003 at 08:33 AM.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #2 (permalink)  
Old 30th March, 2003, 04:27 AM
perc
Guest
 
Posts: n/a

hey im glad you pointed this out .. ive been ripping my hair out for months now trying to get a good oc on this chip.. i must say that i can run the 200/400 fsb all day long with out a volt mod but thats all nothing higher than 205/410 will do jack it just crashes.. im seriosly thinking about that mod and cutting my l3 and seeing what and if that helps? i sure hope the new 400fsb chip isnt like this? oh well atleast you opened my eyes to this chip and i can just quit trying and deal with it till i can afford another chip..


thanks perc.
__________________
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #3 (permalink)  
Old 30th March, 2003, 05:16 AM
muzz's Avatar
Member
 
Join Date: November 2002
Posts: 1,018

perc you have the 8RDA..... have you pulled the JCLK jumper off yet?
That will set the default multi to 100mhz............ which is what the L12 mod does...... I'm NOT CERTAIN it is the EXACT same thing, but it seems to work for alot of folks that have 333fsb chips on the 8RDA/+.
The L3 is for upper multi's on the 2600+/333 chip... I've done that 1.
__________________
All MODS are done AT YOUR OWN RISK.........plan accordingly..

Fat, Drunk and Stupid is no way to go through life.........

I have a pool and a pond........ pond would be good for you though.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #4 (permalink)  
Old 30th March, 2003, 07:42 AM
perc
Guest
 
Posts: n/a

well actualy i have two problems.. one being that anything above 12.5 wont work and the other being that i cant go any higher than 200/400 on the fsb. both causeing me to only get around a 150/200 mhz oc at most? I seem to get the best performance out of the fsb being at 200/400 but then my oc is only 2200 mhz. i dont know maybe i expect too much? any help or tricks or even just links are fine with me. i cant believe what a "POS" this chip turned into when it comes to anything but its stock speed! my friends 2400 xp is getting better ocs than this and hes smashing me!! oh well hey! Its not like this thing is slow or anything its just that I cant get any extra out of it and that sucks cause i built this rig to oc not to run at stock speeds. I should of done more reading on this chip instead of letting the BIGER fsb blind me in the way it did! well hey thx for you responce it just confirms what i need to do and why i dont want another one of these 333 fsb chips....

thx man.... perc.
__________________
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #5 (permalink)  
Old 30th March, 2003, 11:42 AM
Liquid3D's Avatar
Member
 
Join Date: December 2002
Location: Newport, RI SURF
Posts: 374

Before you make any permanent changes to your chip, try the (100/133)  Jumper suggestion first. That's sounds like an excellent idea. As for the "upper multipliers," below is a close up of one VERSION of the "wire-trick", which should open all multipliers from 4x to 24x;

And here's link to the L3 method mentioned in the caption below. This ONLY requires conductive ink, because the 5th L3 bridge contact (pit) can be circumvented by "drawing" the trace around the outside of "pit row", (pardon the pun); http://www.overclockers.com.au/article.php?id=118669
Attached Thumbnails
Has cause been determined for 333FSB CPU's inability to attain high FSB OC's?-vsspinunlock.gif  
__________________
(1.) 2.4C@3.6GHz 300FSB, SLK900U/Vantec Tornado 92mm/119CFM, Abit IS7-E, 1GB OCZ GolD Dual Channel 3700, Tyan G9700 Pro, PCPower&Cooling 400W
(2.) JIUHB1700, Prometeia, Soltek 75FRN2-L, 1GB OCZ EL-PC3700, Leadtek Winfast A250, ANTEC 480W


Last edited by Liquid3D; 30th March, 2003 at 11:54 AM.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #6 (permalink)  
Old 30th March, 2003, 11:42 AM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

Re: Has cause been determined for 333FSB CPU's inability to attain high FSB OC's?

Quote:
Originally posted by Liquid3D
It occurs to me the immaturity of the AMD 166FSB processor, hasn't revealed enough of these circumstances to warrant AMD technicians and motherboard chipset makers into a collaborative research project regarding the issue, and finding a proper fix. Perhaps AMD's attitude is, the small percentage of overclockers/enthusiasts whom push their processors to these "insane" FSB speeds doesn't justify the cost of such research? Or perhaps AMD feels it's mainboard chipset makers responsibility (being the anomally is in the FSB indicative of a mainboard chipset design) OR a juxtiposition of this attitude amoung mainbiard chipset makers, who feel the design flaw resides purely in the processor design. Either way as DDR400 is supported in KT400, and nForce2 chipsets I feel it's unfair to expect end-users to void their processor's warranties to exploit a "standard" feature mainboard makers clearly used as a selling/marketing point.
AMD specify 166MHz as the maximum clock speed for the Athlon XP Model 10 processor in their data sheets (document number 26237B). AMD have quite clearly ensured that their processors meet all published specifications. What is there to fix?

As far as KT400 and nForce2 chipsets go, one does not need to run the FSB at 200MHz to have a ~200MHz memory bus. Hence, it's quite possible to have a 166MHz FSB, with a ~200MHz memory clock. Now you have an Athlon XP model 10 processor running with DDR400. Both sets of products have met their design criteria. The motherboards csupport DDR400, as claimed, and the processor runs at the specified maximum FSB.

What is there to research? Why a chip when run beyond it's maximum specifications fails to operate correctly? Well, that's simple, it's run outside it's specifications. That's a complete non-issue.

Áedán
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #7 (permalink)  
Old 30th March, 2003, 12:06 PM
Liquid3D's Avatar
Member
 
Join Date: December 2002
Location: Newport, RI SURF
Posts: 374

Re: Re: Has cause been determined for 333FSB CPU's inability to attain high FSB OC's?

Quote:
Originally posted by Áedán
What is there to research? Why a chip when run beyond it's maximum specifications fails to operate correctly? Well, that's simple, it's run outside it's specifications. That's a complete non-issue.
Áedán
I see I have an adversary, to several of my theories. Although at times I feel your conclusions are terse, I won't take it personally, as you don't know me personally. I do welcome the intellectual stimulation. You certainly keep me on my toes, and for this I thank you. I do concur with your point of the CPU being pushed beyond it's design spec's, in so far as the memory frequency capability, you got lucky on that one (just kidding).  PS: Don't let the avatar mislead you.
__________________
(1.) 2.4C@3.6GHz 300FSB, SLK900U/Vantec Tornado 92mm/119CFM, Abit IS7-E, 1GB OCZ GolD Dual Channel 3700, Tyan G9700 Pro, PCPower&Cooling 400W
(2.) JIUHB1700, Prometeia, Soltek 75FRN2-L, 1GB OCZ EL-PC3700, Leadtek Winfast A250, ANTEC 480W

Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #8 (permalink)  
Old 30th March, 2003, 01:04 PM
Holst's Avatar
Member
 
Join Date: September 2001
Location: Leics UK
Posts: 4,528

Aidan is right, I very much doubt AMD will do anything about this issue.

As for mainboard manufacturers several allready have the ability to lock at 100fsb (like 8rda+) this does seem to help.
But I dont think anybody will be going to too much trouble looking into this problem.

What I still dont unerstand is WHY this happens, nobody has ofered any sort of credible explanation as to why at CPU set to 333bus overclocks worse than a 266model...
Answer that and the rest should be allot easyer to work out...
__________________
No longer Epox Tech.

Best of luck in the future all my friends.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #9 (permalink)  
Old 30th March, 2003, 01:41 PM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

Quote:
Originally posted by Holst
What I still dont unerstand is WHY this happens, nobody has ofered any sort of credible explanation as to why at CPU set to 333bus overclocks worse than a 266model...
The cynical side of me says AMD don't want to undercut a 400MHz bus version if/when it comes out...

The less cynical side of me wonders if AMD have changed the I/O drivers on the chip, and are still fine tuning them.

Áedán
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #10 (permalink)  
Old 30th March, 2003, 01:44 PM
Holst's Avatar
Member
 
Join Date: September 2001
Location: Leics UK
Posts: 4,528

This dosent feel deliberate to me, if they wanter to cartail overclocking why not just lock the multiplyer and do it propperly...
__________________
No longer Epox Tech.

Best of luck in the future all my friends.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #11 (permalink)  
Old 30th March, 2003, 01:54 PM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

Quote:
Originally posted by Liquid3D
Although at times I feel your conclusions are terse, I won't take it personally, as you don't know me personally.
The answer was terse, mainly due to the way the original post was worded.

Statements like "It occurs to me the immaturity of the AMD 166FSB processor, hasn't revealed enough of these circumstances to warrant AMD technicians and motherboard chipset makers into a collaborative research project regarding the issue, and finding a proper fix. " come across to me as overly hostile. The statement is clearly laying blame where there is no blame to place.

If the wording had been along the lines of Holst's "What I still dont unerstand is WHY this happens, nobody has ofered any sort of credible explanation as to why at CPU set to 333bus overclocks worse than a 266model...", I would have offered an answer worded differently. However, if someone wades in with all guns blazing when the issue could be dealt with in a less confrontational style, they're more likely to get short change from me. It's not that I won't answer to the best of my technical ability (unless I'm seriously annoyed, which is fairly rare). I'm not for withholding information.

Áedán
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #12 (permalink)  
Old 30th March, 2003, 01:58 PM
Chief Systems Administrator
 
Join Date: September 2001
Location: Europe
Posts: 13,075

Quote:
Originally posted by Holst
This dosent feel deliberate to me, if they wanter to cartail overclocking why not just lock the multiplyer and do it propperly...
Just had a thought... As I understand things, running a higher FSB requires more power in order to drive the I/O pins into their correct states. It may be that AMD did a redesign of the I/O buffers to try and reduce their power requirements, to keep the chip within some kind of power budget.

If that was true, it might indicate that AMD's looking at FSBs far above the 200MHz clock, somewhere in the nearer future.

Áedán
__________________
Any views, thoughts and opinions are entirely my own. They don't necessarily represent those of my employer (BlackBerry).
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #13 (permalink)  
Old 30th March, 2003, 02:05 PM
perc
Guest
 
Posts: n/a

liquid3d

this wire trick you speak of? i have seen this before and have considered this option. this will infact unlock my cpu's 13x and up multipliers? i think ill have a go at it today. ill get back to you on this and let you know if it did infact help me.
thx man perc
__________________
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #14 (permalink)  
Old 30th March, 2003, 03:40 PM
Holst's Avatar
Member
 
Join Date: September 2001
Location: Leics UK
Posts: 4,528

I dont think so perc,

You will have to cut the fith L-3 bridge to get more than 13X mltiplyers.
__________________
No longer Epox Tech.

Best of luck in the future all my friends.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #15 (permalink)  
Old 30th March, 2003, 04:32 PM
surlyjoe's Avatar
AOA Staff
 
Join Date: September 2001
Location: caliland
Posts: 2,723

Quote:
Originally posted by perc
liquid3d

this wire trick you speak of? i have seen this before and have considered this option. this will infact unlock my cpu's 13x and up multipliers? i think ill have a go at it today. ill get back to you on this and let you know if it did infact help me.
thx man perc
the easy way to do the wire trick is just bend it in a U and drop it in the holes on the socket before the chip goes in..
__________________
"Many people die at twenty five and aren't buried until they are seventy five"
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #16 (permalink)  
Old 30th March, 2003, 04:40 PM
Holst's Avatar
Member
 
Join Date: September 2001
Location: Leics UK
Posts: 4,528

But it will only let you use over 12.5, not under 13X.

On a CPu with a default multi under 12.5 L3-5 is closed (as if the wire trick was allready done)
To get over 12.5 you need to remove the wire or cut the bridge.

On KT400 and Nforce2 chips with defaults over 12.5 are fully unlocked, but chips below 13 as default cannot use more than 12.5 without mods..
__________________
No longer Epox Tech.

Best of luck in the future all my friends.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #17 (permalink)  
Old 30th March, 2003, 07:57 PM
perc
Guest
 
Posts: n/a

But it will only let you use over 12.5, not under 13X.

On a CPu with a default multi under 12.5 L3-5 is closed (as if the wire trick was allready done)
To get over 12.5 you need to remove the wire or cut the bridge
------------------------------------------------------------------------------------------------------------------


ok well i can use 12.5 and all the way down to 1x right now already i just cant use anything above 12.5 so once i do the wire trick i wont loose my 12.5x and lower will I?? i want to be able to use from 1x all the way to the max multiplier is that possible just by doing the wire trick? If this is the case then thank you that is the answer i was looking to hear

ths guys perc.
__________________
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #18 (permalink)  
Old 30th March, 2003, 08:11 PM
perc
Guest
 
Posts: n/a

hey guys i got a quik question? since im having such a hard time with this 2600xp you think i should just trade my old man for his 2400xp? i dont know how he feels about this but i think i may be able to talk him into it? i mean he dont oc anyway. would that be a better option or am i going backwards here?

thx perc.
__________________
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #19 (permalink)  
Old 30th March, 2003, 11:35 PM
Holst's Avatar
Member
 
Join Date: September 2001
Location: Leics UK
Posts: 4,528

Quote:
Originally posted by perc
But it will only let you use over 12.5, not under 13X.

On a CPu with a default multi under 12.5 L3-5 is closed (as if the wire trick was allready done)
To get over 12.5 you need to remove the wire or cut the bridge
------------------------------------------------------------------------------------------------------------------


ok well i can use 12.5 and all the way down to 1x right now already i just cant use anything above 12.5 so once i do the wire trick i wont loose my 12.5x and lower will I?? i want to be able to use from 1x all the way to the max multiplier is that possible just by doing the wire trick? If this is the case then thank you that is the answer i was looking to hear

ths guys perc.
No, at the moment your fith lever 3 bridge is closed, and you only have <13x
Doing the wire trick wont make any difference at all to what multiplyers you can get.

to fully unlock you need to breat the fith level 3 bridge, breaking bridges is a good way to kill your CPU.

I think you will do better trying to get your FSB up.
__________________
No longer Epox Tech.

Best of luck in the future all my friends.
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
  #20 (permalink)  
Old 31st March, 2003, 03:26 AM
perc
Guest
 
Posts: n/a

Holst''''

No, at the moment your fith lever 3 bridge is closed, and you only have <13x
---------------------------------------------------------------------------------------------------------

I have all my multipliers including 12.5 and down right now and your stament "you only have <13x" is throwing me off?? i really want to try this wire trick but now im all confused!

thx perc.
__________________
Digg this Post!Add Post to del.icio.usBookmark Post in TechnoratiFurl this Post!
Reply With Quote
Reply



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
witch Epox board oc's a xp the best? brian770 EPoX MotherBoards 4 29th August, 2004 09:01 PM
8rda3+ oc's worse than 8k3a+!? saw666 EPoX MotherBoards 7 17th June, 2003 12:23 AM
Strange comment on OC's front page... mookydooky Random Nonsense! 22 20th December, 2002 12:05 AM
IDE RAID 0 and FSB OC's SteveI General Hardware Discussion 17 24th September, 2002 01:13 PM


All times are GMT +1. The time now is 09:17 PM.


Copyright ©2001 - 2010, AOA Forums
Don't Click Here Don't Click Here Either

Search Engine Friendly URLs by vBSEO 3.3.0