Why wasn't the 386sx CPU developed as a drop-in replacement for the 286?

[Quagmire]

Link please.

A quote without the source is worth dilly squat......

When did the 386 overtake the 286?

The Intel 80386 was released in 1985, but was initially expensive, and took a long time to fully displace the earlier 80286 from the market; subjectively, I remember significant numbers of 286 mach...

retrocomputing.stackexchange.com

BTW this is a discussion not a court case.

 

[Caluser2000]

When did the 386 overtake the 286?

The Intel 80386 was released in 1985, but was initially expensive, and took a long time to fully displace the earlier 80286 from the market; subjectively, I remember significant numbers of 286 mach...

retrocomputing.stackexchange.com

BTW this is a discussion not a court case.

I stand by my statement your honor.......

AO you are guilty as charged

 

[Quagmire]

I stand by my statement your honor.......

AO you are guilty as charged

It is what it is and you can Google just like I did. So, what's your beef? You got what you asked for - a reference.

 

[yuhong]

The 80286 existed in CLCC, PLCC, PGA and QFP . Initial shipments were made as CLCC--to my knowledge the 386 never used that packaging.

Which also reminds me I should ask why the 387DX was packaged in 68-pin PGA instead of CLCC/PLCC

 

[Chuck(G)]

My recollection was that the '387 was a late addition. Witness the number of early 386DX boards using the 40 pin 80287 because the 387 wasn't ripe for shipping. So it makes sense that the packaging for the 387 was limited. Do any boards exist where the 386 was a PLCC with the 387 being PGA? I don't recall any offhand.

 

[krebizfan]

There were a number of motherboards designed with a QFP mounting for an AMD 386, a mounting for PGA socket for a 386, and a PGA 387 socket. http://www.dosdays.co.uk/topics/images/mobos/ali386_486dlc/other_1.jpg is an example I could find quickly.

 

[Gary C]

I bought a 386SXnow unit for work to see if I could get it to work with the intel 286 multibus. While I did not get the cache running, it worked fine right up until I tried to access I/O.

I cant quite remember the details now, but the machine hung until I pulled down the interrupt as it was not getting it properly from the I/O interface.

Had to put in in my AT instead.

 

[Chuck(G)]

There were a number of motherboards designed with a QFP mounting for an AMD 386, a mounting for PGA socket for a 386, and a PGA 387 socket. http://www.dosdays.co.uk/topics/images/mobos/ali386_486dlc/other_1.jpg is an example I could find quickly.

Sure, because the AMD 386 was cheap--I have one such board. But a PLCC 386 and PGA 387? I don't think I've run into that.

 

[yuhong]

Sure, because the AMD 386 was cheap--I have one such board. But a PLCC 386 and PGA 387? I don't think I've run into that.

I don't think the 386 would fit in a PLCC package anyway.

 

[Chuck(G)]

You have a point. I have some 84 pin PLCCs, but that's still pretty far from a 132 pin count. I guess I was thinking of the 80286 varieties. Even the SX had too many pins-100.

 

[yuhong]

You have a point. I have some 84 pin PLCCs, but that's still pretty far from a 132 pin count. I guess I was thinking of the 80286 varieties. Even the SX had too many pins-100.

Which is why I focused on the 387DX, which the average user is more likely to touch than the 386DX anyway.