AMD Joining Intel with PIN/Socket Changes

[DCB_AU]

Here we go again. AMD changing pin and socket configerations just like Intel.

EG. If your not very financial at the moment and you instal a 2400Cel now so that you can upgrade later to a P4 when finances allow, you can't do it without upgrading to a new M/board.

Socket 478 are being discontinued.

Laws should be introduced just like the car market so that components must be available-produced for a reasonable number of years!

Sure component changes must happen as technology changes but supply superceded parts for a reasonable amount of time afterwards.

<font color=red>DCB</font color=red><font color=white>_</font color=white><font color=blue>AU</font color=blue>

 

[endyen]

My wholesaler will still sell me a K6-2 500. There has not been a board made since then, that I cant get a chip for. Sure, the price may seem way out of line, but most of that is storeage, and original price.
Bothe companies have gone through a number of configurations, but for the most part, they try to maintain upgradeability. Sometimes, improvements, and new tech, just cant be supported on an old socket.

 

[RichPLS]

Why don't they just quite parleying and merge to the same slit, guys we do it all the time!

_____________________________________________
<font color=red> And the sign says "You got to have a membership card to get inside" Huh
So I got me a pen and paper And I made up my own little sign </font color=red>

 

[Crashman]

AMD lulled you into a false sense of security by FORCING YOU TO BUY NEW BOARDS with the SAME SOCKET. So their compatability with old platforms has never been better than Intel's.

Think about it a minute, K6 worked on a Pentium board, K6-2 required a new board with lower voltage and usually higher bus speeds. Intel went to Slot 1 when AMD went to Super 7. Intel went to Socket 370 just after AMD went to Slot A. But here's the killer: Socket 370 processors worked in Slot 1 boards via an adapter, Slot A processors never worked in Super 7 boards.

Intel kept Socket 370 while AMD went to Socket 462, that means AMD went one more socket at this point. And no, you couldn't get an adapter to Slot A boards. Intel released Socket 423, after AMD released a new bus that required new boards (266). Intel released Socket 478, then AMD released a new bus speed that required new boards again (333). Most Socket 478 boards supported 533 bus. Socket 478 adapters worked on Socket 423 boards.

AMD released another new bus speed (400), Intel released a new bus speed (800).

Forget the fact that sockets stayed the same, new processors usually didn't work in older AMD boards, so you still had to change boards. And unlike Intel, you couldn't just get adapters for AMD.

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>

 

[darko21]

If I remember right Intel patented the slot1 to make thing more difficult for amd. thus super7 slot1, Making a slot A work work in a super7? Could that even be possible. How could a slot1 adapt to pentium socket7 or 5? Might be technicaly possible but not realistic form a design point of view. However powerleap or whoever should have at least offerd an adapter to alow xp's or durons fit in a slot A. Guess that is how it goes when you only have 15% market share.

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.

 

[Crashman]

Just accounting for the specification change, Intel went to Slot 1 when AMD went to Super 7, Intel went to Socket 370 shortly after AMD went to Slot A, but you could adapt a 370 processor to Slot 1, while you couldn't adapt (obviously) a Slot A CPU to Super 7. Keeping everything in a timeframe here. You couldn't adapt 462 to Slot A, but you couldn't adapter 423 to Socket 370 either. Timeframes, like I said.

Then again, 423=Williamette, don't know who would want one knowing anything.

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>

 

[endyen]

Yah, dont mention socket 8, it didn't realy count. Which board is it that tulatans work with? Please also dont mention the Intel chipset that ran so many super 7 boards.
Nor should we talk about how it was Intel's decision to go proprietary anyhow. And it wasnt like people had to change boards to get out of the RD-Rimed again fiasco. Sure, you could OC the FSB on a lot of boards so they could run 133, but it wasn't recommended, so most people didn't do it. Most of these changes were made to give better performance. The one that rancles, was Intel's decision to go proprietary.

 

[Crashman]

Socket 8, Slot-2, etc are server interfaces, not in this market.

Tualatin support required an adapter, but Tualatins were released after the P4 and were simply a diversion from the main processor line (that is, they don't follow the time scale).

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>

 

[endyen]

My 233 on socket 8 was not a server. AFAIK they had originally planned on using s8 for P11. Then they decided to go propriatary, and came up with some lame excuse about cache.

 

[Crashman]

Most Socket 8's went into workstations. None that I know of went into PC's, except business PC's which are actually low end workstations. Of course if you built your own...

The price of Pentium Pros eliminated them from the mass market. The reason WAS the cache. Intel decided to go with 1/2 speed cache, off die, so they DID need a card for that.

They could have made the card fit into Socket 8 I'm sure. That would have been nearly ideal because the same socket could have supported PIII's and Xeons, with full speed cache AND L3 cache for Xeons. But Intel was trying to push the price of server/workstation parts upward, hence the separation. And that separation DID work for them.

Of course we would have been left in the P-Pro erra of VRM's too, you'd have to buy an expensive VRM for your 2.0v CPU since the standard Pro VRM was 3.3v. And every time you changed the CPU core, you'd be buying another VRM. Unless someone came up with jumper-adjustable VRM's.

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>