In Sander's latest article, he argued that the dreaded multiplier lock should be located at the CPU core instead of the PCB. He seemed pretty convinced of that. My question is, what do you guys think and if you're still not convinced, what will it take to convince you? Look forward to hear your feedback!
MyDesktop / internet.com Corp
This is what a CPU looks like when it has had the original 3.0 multiplier lock changed to 4.5 multiplier lock. I got this re-marked PII-450 in February and found this spider under the hood. The store has since replaced it with a retail PII-450 which I am now overclocking to 558 MHz @ 2.1 volts.
If anyone has read his articles on supercooling, then you noticed how in the results page, he shows screenshots of a program reporting his multiplier, bus speed, and cpu speed. One of the multipliers is 8.5, and he also used 8. I've e-mailed him asking how he did this and have gotten no response. Now I read this article basically saying he doesn't know how to do it. Well, does he know or not? And how did he get the multiplier to 8?
The multiplier is 'burned/programed'into the chip when it is manufactured... There are a few un-locked chips around, but they were made that way for testing purposes by Intel..
Remember, 'if it sounds too good to be true, it is'.. Why would a 'psychic' charge YOU $20 bucks for some 'winning' lottery numbers..
If it were possible to unlock intel chips, someone would be doing it, we would all know about it, and if it were me, I would be buying a lot of Intel chips and stock..
jjr512, the multiplier you saw may have been one of those rare unlocked chips, or another explanation is, it would be relatively easy to doctor a screen shot of a 'CPUID' panel.. (A lot easier than unlocking the multiplier on an Intel chip...)
[This message has been edited by Diogenes (edited 09-30-99).]
The multiplier is not burned into the chip, is is held there in what I assume is ROM-like memory. It is set by activating the chip
s RESET command, and while it is active, sending certain signals to the chip. The problem is, no one but Intel knows what those signals are or how to apply them.
As for it being a hoax, I don't think that's the case here. Sander Sassen works for Hardware Central and has a lot of articles here. The one I am referring to, Supercool to a Gigahertz, is featured prominently in the Tutorials section, but you can also click here: http://www.hardwarecentral.com/hardw...torials/718/1/ .
I asked him how he changed the multiplier, and his answer was, I admit, a little unbelievable. He said that the chips they used were PIII/500s, and there was a way to unlock the multiplier by drilling into the CPU core in a very specific place, and it was only possible with ultra-high tech (and expensive) drilling equipment. Now I realize how unbelievable this sounds. I mean, for a drill bit to go through the core casing, it would need to have a certain amount of strength that it couldn't be ultra-tiny, but even if it were a half-mm diameter, that would still go through thousands of transistors. And even if you did take it out, doesn't that circuitry need to be there for the chip to function? On the other hand, I can't think of any reason why he would lie about that. I mean, he is a respected contributor to this site. If he doesn't want people to know how to do it, it would have been easier to just say "I won't tell you".
And to Mr. Sassen, if you read this, could you please post some pictures of your drilled CPU's?
Hello,In your articles about supercooling, it looks like you can adjust the multiplier in your CPU's. I think you've gone up to 8. How did you
do this? I thought they were locked. I've asked you this before, but haven't heard back. Please let me know!Thanks,
Justin J. Rebbert, Mall in Columbia
Sorry for not getting back before, but I gets tons of email everyday and I usually try to answer all of them, but yours must have
slipped through. The lock cannot be easily defeated, it requires drilling into the cpu-core with a specific batch of P3/500's. I
have not yet found an easy way of reproducing it, also it requires very techy equipment (with $$$K price tags).
Email : firstname.lastname@example.org
Siteleader at HardwareCentral
Visit us at : http://www.hardwarecentral.com
While I tend to think the article itself is more or less accurate, I think he's having you on with the drilling comment. It's quite possible they managed to get hold of an unlocked PIII.
I would also not be surprised if there were in fact a way of defeating the lock without opening the core itself. This is the kind of information you might not want to publish though. First you might well get some negative attention from Intel, but perhaps more importantly you wouldn't really want to help out the remarkers would you? That's not good for any of us.
This is the way things have always been. A good hacker will at some point come to know certain things they might not in fact wish to share too freely...
I'm just frustrated because I know that my particular P-III/500 is from the same batch that Intel used for some 600s, simply by changing the multiplier. So, but for that, I could easily have a 600. I can get 600 by overclocking the FSB, but then again, if I could change the multiplier to 6, then I could overclock to past 600, like maybe at least 672. But the point is, it didn't cost Intel any additional money to set the multiplier at 6 instead of 5 at the time of manufacturing, so I feel that I, as well as anyone else with the same batch of chips, should be able to 'upgrade' them, either for free, or I'd even be willing to pay a small fee for it (like $50).
Well, that's my 2¢.
jjr512, I really respect your input to the posts here, you have brought up some very interesting points. I was being a bit simplistic when I used the term 'burned'.. I am not an engineer, but I agree the process is like setting up a PROM. I have heard of that referred to as 'burning', and I am reasonably sure the process is irreversible.. I have heard of the possibility of adding some wired logic to the P3 'card' to change the multiplier, but it probably is not worth the time and trouble. I still submit, if it were economicaly feasible to 'unlock' the multiplier, it would be done by now.. It's like, if you have enough money to turn a Yugo into a Ferrari, you just buy a Ferrari in the first place..
Anyway, how do you clip off a pin that's printed on the circuit board? Or did you mean from the actual CPU core? How did you here about this, and can you prove that it works? It's not that I don't believe you, I just want to know first.