• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

It’s over. All AMD CPUs except for the expensive FX will be multiplier locked.

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
I think what Pikachu_Mommy is trying to get at (while also tossing the flame-bait about, and innaproriate) is that *some* nForce 2 boards aren't all that well made.

Take my ABIT NF7-S for example. Great board, I love it - I have nothing against it, or AMD, and am *EXTREMELY* happy with the results I have gotten with my AMD/nForce 2 system. That's my little anti-flame bit for all users of the same/similar hardware.

My board and XP 1800+ will pass memtest86, Prime95, SuperPi, HotCPUTesterPro, toast, etc. for 24 hours plus at a 26XX MHz 233 MHz FSB -> with my OCZ PC3500 EL RAM. So, my RAM isn't the problem, just like Pikachu_Mommy's RAM isn't his problem.

In 3DMark, and 3D aps I get freezing at anything above a 211 MHz FSB. So what is the problem? My RAM? My processor? OR maybe it is my motherboard - the AMD platform I am using. At 211 X 12.5 (2638 MHz) my system is totally, and completely stable - but the platform is limiting my FSB overclocking ventures -> if my processor was multiplier locked, I couldn't run my processor at the 2638 MHz I run it at daily. The AMD platform is simply not suited for FSB overclocking - nevermind you who have gotten to 230+ FSB stably - the large majority of us cannot get our motherboards stable past 210 FSB - and it has nothing to do with "overclocking skills".

Please don't tell me that I am an unskilled overclocker or that I don't know what I am doing, as you disrespectably did to Pikachu_Mommy, because both of us know what we are doing here, and what we are talking about.

Pikachu_Mommy is wrong to bait you guys, but you are even more wrong to take his bait, and provoke him even more - when what he is saying is very valud information.

I am disgusted by the way some of you have been talking to each other on the first page of this thread.

c627627 - I find it interesting that the Athlon 64's will be locked, while only the FX chips will be unlocked. I would have thought that AMD would keep all of the 64-bit desktop processors unlocked. In the long term I hope that the AMD platform improves, and begins to develop to allow for more reliable FSB headroom - which would make the lockededness of the new processor unimportant, and no roadblock to our overclocking ventures - we overclocked with locked AMDs for a long time before this, and we can again - if the platform can grow to support it.
 
Pikachu_Mommy said:

I need a well matured P4 platform..

youre calling the P4 well matured....? good lord son, the P4 has had many architecture changes since its debut. the K7s core architecture has been the same since its debut....now which product is mature? have you ever seen AMD change the PGA for the athlon? i know ive seen intel change the PGA for the P4....along with a few other things to boot. if its FSB youre after, get a barton....because tbreds arent FSB mongers. im lucky mine will do 218FSB stable. now bartons on the other hand, with good mem and a good mobo, 220+ shouldnt really be a problem.
 
Ah, yes...but that's what the L12 mod is for. :)

You know, I wouldn't be surprised if more and more methods for high FSBs came out. Unlocked, you don't HAVE to get a high FSB for a good overclock. But if the only option is FSB overclocking, then I'm sure people would find a way to get better results.
 
the best results would be in the hopes of a 1600+ DLT3C because of its 10.5x multi....but i dont think they make one of those. or maybe they could do a 2400+ barton as well with a multi of 10.5x. stuff like that wouldnt be so bad to have locked, if they actually made such chips. we can only hope lol.
 
or an unlocked applebred duron 1400 i think the lowest thorton core is the 2000 so i guess that would be out....
 
Is there a way to do a bridge or pin mod to change the multiplier? That still gives more flexibility to AMD users, provided you're willing to tweak a bit.
 
unless I mesread this entire thread, all AMD will be doing is what they did back in the tbird era where they just didne't cut the bridges to lock the chip. back then we'd get out our BRAINS and figure out a way of fixing it instead of flaming back and forth on a matter totally unrelated to locked AMD chips.

now does anyone know if they are moving away from the bridge setup on the XPs and MPs, or are they keeping that scheme the same? this is the first theing we need to know before we can start screaming about the end of the world.

the moral: We found a way around locked multis before, we can do it again. :)
 
Extensive tests were done, one of the forum members (Hoot) even disecting a working locked Barton. There is nothing that can be done to bridges or pins that can change the locked multiplier because they are not connected to the circuitry. They cannot be unlocked by any of us.
 
Deathknight said:
Pikachu_Mommy try and calm down a little bit man. You obviously have a deep seeded hatred of AMD for some reason, probably due to some frustrating OCing experiences with them, but the fact is there is a great # of people that have had great success with them. Its fine to have your opinion, but you are coming off quiet abrasive when there is no need.

ok


If your system cannot do 160fsb, there is something seriously wrong with your setup. You got a bad piece of hardware, or you got something configured wrong. Looking at the polls I have seen on these very forums more than half of the people here fall into the AMD camp and I would think these forums would be flooded with complaints if nobody could game on their system without going down to 100fsb...
umm.. i like this comment. :D
 
Sir Barton said:


if its FSB youre after, get a barton....because tbreds arent FSB mongers. im lucky mine will do 218FSB stable. now bartons on the other hand, with good mem and a good mobo, 220+ shouldnt really be a problem.

umm im da one was gonna say that... not coz i don't know but becoz i haven't talked about barton chips yet.. of course.. with da barton 2500+, I did gaming fine at around 230FSB with same ram, same mobo(the one that fails 160fsb in games with Tbred 133fsb chip)... but my favorite oc'ing chip is the Tbred 133fsb chip.. and this tbred chip does just fine at 220 FSB with the KT600 board in games.. like that noobie Cuda keeps saying my ram is causing it then why does it run so good with my KT600@ 220FSB and why did it run so good at 220FSB in one of the Abit NF7 v2.0 boards and did 230FSB with the very first NF7 v2.0 I had... and 220FSB with the v1.2 and 1.1s.. and 217FSB with v1.0? and how did it repetetively pass Memtest86 at 225FSB? Still my ram is holding it back? that is the worst response I have ever gotten in my entire life....
 
Last edited:
Pikachu_Mommy said:


umm im da one was gonna say that... not coz i don't know but becoz i haven't talked about barton chips yet.. of course.. with da barton 2500+, I did gaming fine at around 230FSB with same ram, same mobo(the one that fails 160fsb in games with Tbred 133fsb chip)... but my favorite oc'ing chip is the Tbred 133fsb chip.. and this tbred chip does just fine at 220 FSB with the KT600 board in games.. like that noobie Cuda keeps saying my ram is causing it then why does it run so good with my KT600@ 220FSB and why did it run so good at 220FSB in one of the Abit NF7 v2.0 boards and did 230FSB with the very first NF7 v2.0 I had... and 220FSB with the v1.2 and 1.1s.. and 217FSB with v1.0? and how did it repetetively pass Memtest86 at 225FSB? Still my ram is holding it back? that is the worst response I have ever gotten in my entire life....

i doubt its your ram holding you back....could just be the chip itself. when i get my BH-5 going on im going to L12 mod my CPU and see what kind of FSB i can get.
 
I just look at this as an opurtunity to experiment and learn new OCing techniques for the future batches of chips!
 
johan851 said:
Have you tried the L12 mod? TBred's don't do well without it.
of course.

no effect tho.. Gained 1mhz FSB in board A while board B gained 3mhz FSB and board C gave zero gain. that was with both 166fsb and 200fsb mod.. no effect on either ones.

Bios10 helped a little bit too, on the board A. Gave no effect and made it even worse with board B and C.

now.. board A, B, and C is not the only v2.0 boards I had... probly had over 10 of v2.0 boards. I had wonderful experience with v1.0, 1.1, and 1.2s tho.
 
Sir Barton said:


i doubt its your ram holding you back....could just be the chip itself. when i get my BH-5 going on im going to L12 mod my CPU and see what kind of FSB i can get.
good luck yo.. you might get a bit of luck and get some good FSB oc'ing. :)




and to felinusz, hey thankx... for understanding me where I was coming from.. :)
 
Last edited:
It'll just make it harder work. that's a shame, but not the end of the world.

I assembled a computer for my boss. We have a deal through a wholesaler at work. I helped him pick the stuff. A7N8X, 2400+ 512 PC2700, 9200 Radeon.

I hadn't really been paying attention to what a locked multiplier on a 2400+ meant. 133 FSB! On a board where overclocking the FSB means pushing it past 200 MHz!

He'll never know the difference, it is replacing a P200 AT box. He thinks it is great, and for the dollar, it is a very good machine.

But it was hard for me to accept that I had to leave it set where I used to run my old Abit KT7A. Actually, I used to run it up to and a little over 150 FSB.

Some of you guys need to look back a couple of years, and think of all of the gizmo's (Golden finger devices) software (Soft FSB) etc.

Think of the default performance of coming machines!

So lets get to work on finding out how to unlock these coming CPU's.

And thanks for the information c627627, as sad as it is.
 
i think it changes the whole scope of things.

the balance of multi's and FSB made overclocking AMD desirable, not to mention if you killed a chip its cheap to replace it.

but now, with most all desireable XP's gone (cheap, stellar cores, unlocked), if i were to build a new system with OC'ing in mind knowing its gonna be FSB only, i'd seriously consider going Intel myself, and ive never owned an intel powered desktop. (one intel powered laptop)

i know decent overclocks will still happen on locked AMD chips, but its gonna mandate great memory, a board that will run higher FSB's, and if one has to worry about that, they may as well go Intel. its no secret that intel chipsets handle high FSB's better.

AMD had better hope that chipset makers get some FSB headroom soon if they wanna keep the enthusiast market.

i dont blame AMD for trying to motivate people to higher priced chips, but unlocked chips should be at maybe the $150 price point rather than $700 for an FX.

if AMD really thinks people are gonna stick true and pay $700 for the priviledge of unlocked multi's, their totally NUTS.
 
Last edited:
One question for AMD users:

Have you all aplied the registry fix in windows for Athlon?

There is a bug in windows on systems with Athlon CPU and AGP video card. If the bus is heavy loaded there is a chance of a dead lock. The result - the system freezes on a kurnel level.

That's why prime95, folding, memtest, ... don't freeze. When you run a heavy 3D game it freezes after some time when the chance takes its toll. That's why 3Dmark doesn't fail - it runs only a few minutes.

I had problems playing UT2003 - after 2-3 hours, or 10 minutes it freezes. Some times I have played for 10 hours without a freeze.

After I aplied the fix - no freezes since then.

So once again - this fix is not included in any of the patches (service packs).

Here is the link:
http://www.winguides.com/registry/display.php/798/
or here:
http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_756_759^871^2367,00.html
(couldn't find the microsoft link)
 
Last edited:
Pla said:
One question for AMD users:

Have you all aplied the registry fix in windows for Athlon?

There is a bug in windows on systems with Athlon CPU and AGP video card. If the bus is heavy loaded there is a chance of a dead lock. The result - the system freezes on a kurnel level.

That's why prime95, folding, memtest, ... don't freeze. When you run a heavy 3D game it freezes after some time when the chance takes its toll. That's why 3Dmark doesn't fail - it runs only a few minutes.

I had problems playing UT2003 - after 2-3 hours, or 10 minutes it freezes. Some times I have played for 10 hours without a freeze.

After I aplied the fix - no freezes since then.

So once again - this fix is not included in any of the patches (service packs).

Here is the link:
http://www.winguides.com/registry/display.php/798/ (couldn't find the microsoft link)

actually didnt know it existed, but ive never had mine lock gaming.
 
Ah, but what about the Mobile Athlon XPs? If you're still in the market for a desktop chip and you have to have it unlocked, these are the way to go. They have to be unlocked for Powernow! to work. :)

As for the FSB headroom, I wouldn't worry that much about it. Some nForce3 boards are able to hit 300+, and that's with no PCI lock. So, with a properly implemented chipset, good memory ratios, and a PCI lock, there should be no issue about overclocking headroom.

Stop freaking out about this, guys. AMD has been delivering a great product at a great price for a long time, and they'll continue to do so. What does Intel have in the price range of the 2500+? A bloody Celeron. You're still going to have to spend twice as much for a comparable Intel chip, and on top of that, it's going to be locked too. So what are you getting by switching to Intel? Nothing that I can see, especially if you go with S478.
 
Back