ATTN: Intel Northwood Owners...

[PsycoPhreak]

OK, now that I have your attention...

I'm doing some personal research into the P4's, more specific, the northwood cores. What I am trying to learn is how the cpu's are multiplier locked. Now I know on paper the locks are *suposedly" implemented within the core itself, as everything I have read and everything I have been given to read state as such.

Lately I have been second guessing this, although I can be wrong, I figured I'd spend a little time and expend a small amount of effort finding out the answers to the few questions I still have unanswered.

Maybe it has been covered before, I may have missed it, but recently, something odd has been brought to my attention that I didn't notice before. If you were to flip over any northwood, on the bottom of the cpu package you'll find a grid of small pads, some have components (of which I'm not sure what they are, yet ) soldered onto them, some are left completely open.

What I have noticed upon examining one of my P4 northwoods ( in this case, an unused 1.6a ) is that many of these pads remain untouched, as well as a few other 1.6a's I have gotten a chance to see, and what got me thinking is, the "grid" of pads and "small components are setup exactly alike. The links below are to pics of the front and back of my 1.6a, take a look at the pic of the back to understand what i'm refering to.


Front

back

Again, all the 1.6a's I have had a chance to see looked exactly the same.

I yanked out my 2.8c from my gaming rig to have a look ( forgot to get pics ) the "grid" of pads is the same, but the "small components" are placed differently, got me thinking more. Then a friend of mine who loves to brag about things he has told me he got his hands on a "ES" p4...so I asked him about the "pads" on the back, his reply was "what pads?" Come to find out, it appears that ALL of the pads are used by these "small components" on his chip ( according to him ), made me think even more. Now I obviously don't have access to any "engineering samples" so I can't know for sure if they all are like his, but I have seen 2 or 3 pics of suposedly "ES" p4's and they too had every available "pad" on the back used. So now I'm thinking, that with the possible help from people here, especially those who may be fortunate enough to have an "ES" to take a look at the back of your northwoods, regardless of their speed and if at all possible, post a link to a pic so that I may get an idea as to the placement of these"small components". Model/stepping info would also be welcome as it may help in figuring out just what the function of these are and how they *might* play a role in the p4's multi lock.

As I said, I may be completely off base, and all these little pads and components may be for voltage regulation and whatnot, etc etc etc...but how can I know unless I try right?

P.S. correct me if I'm wrong, but Intel mass produces the chips and then configures them ( locks ) according to what their tested speed is, meaning, they check the max speed they can run at and lock them accordingly to prevent passing the stability threshold, this means the chip itself already has to be complete. If so, this would have to be done by some external means, either by blowing out what some refer to as "fusible links" within the core to "lock" the multi, or perhaps by some other external means, thats where my curiousity has gotten the better of me and these "pads" come to mind.

Let me know what you think........

 

[stan03]

i never thought of that, it kinda makes sense, but then again i dont' know much about this. maybe batboy or someone can come in and post some more exact info. i just think that this might have already been thought of. but i may be wrong...

 

[NookieN]

The capacitor packs on the bottom of the CPU are basically there for signal integrity and a little bit of power conditioning. They don't control the multiplier. The 1.6A and 2.8C generally come from very different steppings. Intel often develops a new recipe for those cap packs with each stepping, so it's not unusual for the them to change.

The multipliers are set in polysilicon fuses on the die itself. There's no external wiring, laser cut bridges, or any other wacky stuff like that on modern Intel chips.

 

[PsycoPhreak]

The capacitor packs on the bottom of the CPU are basically there for signal integrity and a little bit of power conditioning. They don't control the multiplier. The 1.6A and 2.8C generally come from very different steppings. Intel often develops a new recipe for those cap packs with each stepping, so it's not unusual for the them to change.

The multipliers are set in polysilicon fuses on the die itself. There's no external wiring, laser cut bridges, or any other wacky stuff like that on modern Intel chips.

Well said, goes along with eveything I have read in whatever tech specs and sheets I have been able to get my hands on, but, like I said, I'm curious to see why the "ES" p4's seem to have all available connections used, where a normal retail, which is locked, has them layed out in specific patterns. thats why I'd like more info on what people have here, and how their's are layed out. I've learned to not believe everything I read or hear, especialy concerning something like this, as Intel, would do/say whatever they had to, to prevent people from buying cheaper cpu's and changing multi's to get a higher performing chip at sometimes half the cost. I've never seen anyone look into this before, and I've learned to not believe all that intel says, on paper, or word of mouth...

Like I said, just doing some personal research, hoping that maybe next time someone tears apart their rig to clean/rearrange..etc, that maybe they'll snap a pic of the bottom and post their model/stepping info. Ya never know where this can go...

 

[Detract]

If you have a lock at the bottom of a M0 2.4C, you will see that all the pads are used and that has a locked multi.

 

[PsycoPhreak]

Interesting, The 2.4c M0 stepping is a good Oc'er, now my curiousity is shifting to the effect that those "caps" have on a potential OC. I have a few paths to follow now, I'm definetly going to keep digging some more, may even attempt shifting around the ones on my 1.6a, wonder if I can get some more off of some used p4's...

Time to go shopping...

P.s. keep the info on the chips rolling in if possible...