//flex table opened by JP

Click to See Complete Forum and Search --> : New Boards Needed For P4 3Ghz


absalom
08-23-2002, 02:08 AM
New Boards Needed For P4 3Ghz

DigiTimes Reports :
Motherboard makers revealed plans to launch two board versions for each of their primary product lines in the fourth quarter in order to accommodate both Intel’s current and new Pentium 4 processors, which will be running at clock speeds over 3GHz. Requiring 70W power supplies, the new P4 processor is incompatible with present motherboards, which only support the 60W standard, board manufacturers said. For the technology update, companies need to change the on-board PWN (pulse width modulation) controller IC design from three-phase mode to four-phase.

In addition, to pack in an increased number of components, board makers have to rearrange the circuit layouts as well. Board makers estimated that these changes will add an additional US$2 to their production costs for each P4 motherboard. The new P4 is expected to hit the market at the end of September or early October. Currently, most companies have already finished the new product design, and some of them have continued to develop boards that will be able to support the latest P4s with the three-phase-mode PWN controller IC, hoping to reduce the costs.

From http://www.digitimes.com/NewsShow/Article.asp?datePublish=2002/08/19&pages=14&seq=84:t

NDD
08-23-2002, 02:15 PM
Too bad for P4 owners, seems like true :(

That's why I'm with AMD :D ;)

Peter M
08-23-2002, 03:50 PM
surprise, surprise ... now how many more times must Intel pull that "hey, buy this, it's future proof ... time passes ... oh oops, we need to change the platform a bit, so please buy our latest future proof thing now" stunt?

Heck, that'll be the FIFTH sub-platform on Pentium-4 alone!

regards, Peter

NDD
08-23-2002, 04:22 PM
More will follow :rolleyes:

gtr
08-23-2002, 07:58 PM
That is the reason i buy AMD. I trust them a lot more than intel. However the 3 ghz sounds very appealing to me. One of these days i'm going to build an intel for myself. I prefer to support amd but the P4's are kicking amd's **** :mad:. Hello 333fsb?

$1500-P4 gamer
08-23-2002, 09:07 PM
Originally posted by Peter Missel
surprise, surprise ... now how many more times must Intel pull that "hey, buy this, it's future proof ... time passes ... oh oops, we need to change the platform a bit, so please buy our latest future proof thing now" stunt?

Heck, that'll be the FIFTH sub-platform on Pentium-4 alone!

regards, Peter
5, how 5? Socket 423-478! Then to 533fsb (which alot of the 400fsb boards will do). Thats 3. Everyone knew about the 423 so really only a fool woulda bought it. So really in all acounts 2 changes. Still too many. But heck, 1.4-3gig is alot of cpu on one board is it not? What other platform went that far? A 1.6gig distance. Think about it. Power is gonna be a issue at a certain point its inevitable.;) :t Plus Intel said the P4 was good for 5gig so far. That means 3gig -5gig. I can see that on one board. Thats a 2gig spread compared to the 1.6gig on the first 478.:t

Kurylo
08-23-2002, 09:34 PM
OMG! 70Watts!!! Our brain takes 20Watts!!!

NDD
08-24-2002, 01:10 AM
What about going from Willamette to Northwood, $1500-P4 gamer ? Not all mobos support it :(

Peter M
08-24-2002, 07:51 AM
Originally posted by $1500-P4 gamer

5, how 5? Socket 423-478! Then to 533fsb (which alot of the 400fsb boards will do). Thats 3.

Sorry, five.

423
478
478 northwood compatible
478 533 MHz
478 70-watt power

It's that kind of subtle changes that is supposed to keep us buying new boards all the time.

It's been like that all the time. Can't remember the details from the 486 days, but here goes for the newer stuff.

Pentium ... first we had Pentium-60 and 66. P75 used a completely different socket, boards in the bin please. That lasted up to P100, because P120 and P133 suddenly needed a multiplyer jumper. OK, buy a new board for this. Got it, want Pentium-166? Oh sorry, we now need a second multiplier jumper. Done, want P-200? Oops, that needs a little more power than P166 ... Hey look, there's Pentium-MMX! No, you can't use it on your two week old boards because we're using another voltage now.

P6 generation ... let's start with Pentium Pro. Hey, we end this incredibly expensive thing at 200 MHz. Here's shiny new Pentium-II. Future proof Slot-1. And blazingly fast 233 MHz to start with ... a little time passes ... here's the new 333 MHz one, oh btw, it uses a different voltage, please upgrade your mainboard. Want 350 MHz instead? Oh oops, sorry we didn't tell you last week that this one is going to use 100 MHz bus. Of course we got new boards for you ... and no, we're not going to make any 66 MHz flavors anymore - not that we couldn't, we donwanna. Got 100 MHz bus now? Well here's shiny new Coppermine core ... no it won't run on your old boards because it uses a different voltage again. Got a new board now? Btw, we're moving away from slots, back to sockets. Oh, you got a Celeron socket board already? No sorry, the new socket looks the same but isn't, you can't use Coppermine in there. Please upgrade ... oh btw, didn't we tell you we'll move to 133 MHz bus? No? Sorry ... got that now. But if you want the new 512-KByte cache Tualatin, you gotta buy a new board because we changed the electrical properties of the bus a bit. btw, we're going to abandon this platform altogether anyway.

Consider the timeframe in which all this happened, and you'll see what I mean.

AMD in comparison has done a lot better, with two changes in socket-7 (to 2.2V and then to 100 MHz bus, the latter not even being mandatory because of the non-locked multipliers), and so far one single relevant change to socket-A, the move to 133 MHz bus ... and if you got an old board, not all is lost because you can still use a Duron in there - provided your mainboard maker didn't intelize their strategy and provided a BIOS upgrade that lets you do it.

regards, Peter

$1500-P4 gamer
08-24-2002, 12:24 PM
Originally posted by ND
What about going from Willamette to Northwood, $1500-P4 gamer ? Not all mobos support it :(

Umm, most do. I havent seen a P4 478 yet that wont, though I know some wont. Lets just say you have to really look to find a 478 non Nortwood support, Heck I got my Asus P4t-e when the 1.5 was still pretty new, and it already had Northwood AND 533fsb support. Just gotta keep your eyes out for this stuff is all.
Now I usually double my cpu mhz on upgrades. I'm 1.5gig now. Have only boughten one intel mobo and can support them all up to the 3gig. Thats not bad at all the way I figure it. Now comes the 3giger. well guess what its my double the mhz upgrade! So new mobo and cpu-oh well. Thing to remember here is were the DIY'ers. Were a rair breed amungst the commen users. They upgrade so late they need a new pc altogether be it AMD or Intel. So I dont think our gripes about the mobo changes hold water when it comes to the commenuser. Intels history proves that. If DIY'ers held more of the market then they would actually care.:(

"AMD in comparison has done a lot better, with two changes in socket-7 (to 2.2V and then to 100 MHz bus, the latter not even being mandatory because of the non-locked multipliers), and so far one single relevant change to socket-A, the move to 133 MHz bus ... and if you got an old board, not all is lost because you can still use a Duron in there - provided your mainboard maker didn't intelize their strategy and provided a BIOS upgrade that lets you do it."

Hmmm, 486 amd to K6. Then most K6 mobo's didnt run K6-2's cause of the vcore issue so ooops change mobo. Then ohh the socket 7 SUCKS cause the agp is junk on it. So then you have the 2.2 version also as you said and then the 100mhz fsb if you were lucky, alot only had 90mhz fsb. Yes it is mandatory, as if changeing the multi doesnt void warrenty just as much as raising the fsb! Then the socket A. Then the socket A 133mhz. And some where in there the change to ddr for XP aswell. Lots of older Socket A's DO have issues running a XP cpu. Then your stuck with Duron. So ditch the mobo.
7 mobo changes, not counting the addition of ddr to XP so then its 8. It all in how you look at it.;)

$1500-P4 gamer
08-24-2002, 12:34 PM
Ooops, 9! Lets us not forget AMD's promised new fsb speed. That will require a 1/5 pci divider most of you dont have. New mobo!:D

Peter M
08-24-2002, 01:04 PM
Originally posted by $1500-P4 gamer



"AMD in comparison has done a lot better, with two changes in socket-7 (to 2.2V and then to 100 MHz bus, the latter not even being mandatory because of the non-locked multipliers), and so far one single relevant change to socket-A, the move to 133 MHz bus ... and if you got an old board, not all is lost because you can still use a Duron in there - provided your mainboard maker didn't intelize their strategy and provided a BIOS upgrade that lets you do it."

Hmmm, 486 amd to K6. Then most K6 mobo's didnt run K6-2's cause of the vcore issue so ooops change mobo. Then ohh the socket 7 SUCKS cause the agp is junk on it. So then you have the 2.2 version also as you said and then the 100mhz fsb if you were lucky, alot only had 90mhz fsb. Yes it is mandatory, as if changeing the multi doesnt void warrenty just as much as raising the fsb! Then the socket A. Then the socket A 133mhz. And some where in there the change to ddr for XP aswell. Lots of older Socket A's DO have issues running a XP cpu. Then your stuck with Duron. So ditch the mobo.
7 mobo changes, not counting the addition of ddr to XP so then its 8. It all in how you look at it.;)

Well I didn't count the 486 at Intel either, so let's leave that out. What remains is the change from 2.9V to 2.2V and the move from 66 to 100 MHz bus throughout the K6 generation ... and as I said, K6 does run on lower bus speeds (with higher multipliers) just fine. Intel's stuff doesn't. And there's been just one not-really-100-MHz board, no truth bending please. And we're not talking graphics card related changes here either. So that makes two and a half iterations for K6.

On K7 platform, we had slot-A to socket-A transition, and one change on the socket. We're on the third iteration of K7 platform now, with the 2nd one still being supported by currently made CPUs. Lack of BIOS updates is not AMD's fault, there's nothing in the hardware that would keep any first-generation socket A board from running an 1.3 GHz Duron.

regards, Peter

NDD
08-24-2002, 02:35 PM
Don't mess with Peter Missel, $1500-P4 gamer, I don't think you can win ;) :)