Why isnt FSB locked?

[metra]

Is there an actual reason?

Cause i could guess that one reason would be that if FSB is locked, then that company would lose a considerable share of the enthusiast market.

However, is there a technical limitation? Any other reasons besides the one i mentioned?

[climbski]

No ... Palomino's are locked. t-breds arn't it's whatever the manufacturer decides they want to do.

[icantgetanyname]

not the mult. .. the fsb.
u can raise the fsb on a pally ... just not much..

[zabomb4163]

long as the multi was low the fsb could go high on palis. Even tbirds can go 200fsb

[bigfoot]

"u can raise the fsb on a pally ... just not much.."

What!!! I have a 1600+ (1.4Ghz) running @ 220x9 in my friends new A7N8X Delux...


And as for what metra was saying, no, knowone can lock the fsb because the fsb is controled by a chip that generates that frequence (usually a chip that is located right beside a crystal w/ the speed of 14.318Mhz) and since the mainboard maker gets to choose that chip and not AMD or Intel, then the fsb will never be locked

unless ofcourse you chose to use an intel mobo that uses a chip that wont suport adjustment!

[asj]

you can only get high FSBs with pallys if you have a good mobo! the mobos that where around when the pallys first came out could not handle it! and to answer the question, Why isnt FSB locked? cos its more fun! hehe!

[metra]

Bigfoot: And Intel cant do anything about it? Although thats what i began thinking myself - that a chipset controls fsb, etc. and not the CPU, doesnt it seem unlikely that Intel cant make a chip that doesnt support varying FSB's? Because of my limited knowledge on this matter, im just talking out of logic.

Can Intel definately not do anything, or would it just be too much of a hassle (changing chip designs, fabs, etc)?

[NookieN]

Originally posted by metra
Can Intel definately not do anything, or would it just be too much of a hassle (changing chip designs, fabs, etc)?

Earlier this year Intel was granted a patent that could be used to prevent overclocking by raising the FSB speed:

http://www.theinquirer.net/?article=8518

Of course, the technology was developed 4 years ago and has not been implemented in any current chips or announced for any future chips.

It seems unlikely that Intel would actually use the technology in their chips, when you consider that the vast majority are sold in OEM computers that can't be overclocked. And since overclocking is becoming somewhat more high-profile, having chips that overclock well is a good prestige factor for a company.

[Zuzzz]

Another factor in this is, in the old days overclocking meant changing the divider. Intel answered with locking the divider. next came the P2 and P3. Oveclocking these chips meant overclocking the whole computer system as at the time you did not have the ability to isolate the FSB of the chip from the AGP/PCI bus. This had its own limitations as you had to to have TOP NOTCH parts to make overclocking possible. The limits of the PCI and AGP bus specificly are not all that friendly when overclocked too far. NOW with Boards like the IC7/IS7 and the Asus P4P800/P4c800 you can lock out the rest of the system. This alone makes such a difference when overclocking as it allows you to only overclock the chip/memory(if you choose). This being said at the time Intel felt it unnecessary to attempt to lock the FSB. I do think however this will change as Intel, I doubt, will allow such EASY and rather extreme overclocks for very much longer considering on AIR most of you with P4 2.4's and 2.6's are easily out clocking their(Intel's) top of the line processors. This is just My 2 cents.

[BlkEnig]

And thank god it will never b locked :)

[zabomb4163]

BlkEnig- never say never ;)

if p4 couldn't overclock....90% of the members on this forum would support amd chips all the way. Each of us would in turn tell others never to buy intel chips and so on. your compeditor having the fastest chip even if its not stock is a bad thing. locking the chips would just be a bad buisness decision.

I suspect the only reason intel locks their multis is to prevent oems from selling overclocked intels.

[bulk88]

If intel puts a ondie oscilator (crystal), that the end of overclocking. I think intel is being stupid by not doing it. Then they can release higher speed cpus and MAKE overclockers buy not matter what the price ($200 more for just 100mhz). Sorry but I am talking from a shareholders point of view.

[bulk88]

Originally posted by zabomb4163
I suspect the only reason intel locks their multis is to prevent oems from selling overclocked intels.

If I were Intel I would put a firmware locked multiplier, so that it will always boot at a certain speed, then if the MB wants it can change it, it can reprogram it, but if you move the procesor to a new board the firmware lock will put it back to stock. The only way around this is with a custom BIOS, and is Dell going to screw with intel (example, enable PAT in 865 chipset), hell no. Another thing to do is to put a error that you need to F1 through if your using multiplier overclocking via disclosure contracts and supply contracts.

Also doesn't intel have the frequency ID program that anyone can use to read WORM in all of their processors. THat WORM can't be changed.

On a side note I figured out how to change the multipliers on a coppermine p3. It is in the resistor in the bottom of the CPU, I hven't tried dealing with them but I think thats how they control the multiplier becasue they are different for each cpu yet the solder pads are always in the same locations.

[NookieN]

Originally posted by bulk88
Also doesn't intel have the frequency ID program that anyone can use to read WORM in all of their processors. THat WORM can't be changed.

The name string is stored in PROM on the chip, so yes it can't be changed. The term WORM refers to an old form of optical storage devices so that's not really an applicable term.

On a side note I figured out how to change the multipliers on a coppermine p3. It is in the resistor in the bottom of the CPU, I hven't tried dealing with them but I think thats how they control the multiplier becasue they are different for each cpu yet the solder pads are always in the same locations.

In the case of P4s, that definitely won't work. The multiplier is fused on-die. As far as I know, the capacitor packs on the bottom of the chips are related to signal integrity and shaping. P3s could be slightly different, but I doubt you would be able to change the multiplier on those either. If you feel like taking a soldering iron to one though, go nuts.

[JaY_III]

Originally posted by bulk88
On a side note I figured out how to change the multipliers on a coppermine p3. It is in the resistor in the bottom of the CPU, I hven't tried dealing with them but I think thats how they control the multiplier becasue they are different for each cpu yet the solder pads are always in the same locations.

now that will just kill your CPU.
think of the L1 bridges on an AMD CPU that selects the multiplier. Now notice how some are cut and that is a locked CPU? Now picture the L1 bridges cut to a locked value on an Intel CPU. But only on the Intel CPU they are placed in the CORE itself.

See the problem? you have no way to connect the CUT bridges and thus a virtually unlockable CPU.

[looktall]

Originally posted by zabomb4163

I suspect the only reason intel locks their multis is to prevent oems from selling overclocked intels.
there's nothing at all to stop OEM's from selling overclocked computers even with the multiplier locked.