• Please review our updated Terms and Rules here

Socket 8 boards with 440LX or BX chipsets.

Mr. Horse

Experienced Member
Joined
Dec 13, 2017
Messages
437
Location
USA:OHIO
OK so for those of you that don't know this, there were socket 8 to slot one converters out there.
See image.

This has left me thinking, if these Slotkets work in 440lx boards and some early 440bx boards, were there any native socket 8 boards that used 440lx or 440bx?
It would be cool to find a socket 8 440lx board with agp, and SD ram support.
 
To the best of my knowledge, the socket 8 slotkets only work with 440FX chipset slot 1 boards, as the later 440BX and kin did not explicitly support the PPro. It might be possible to adapt a 440BX to do so, but by the BX's time Pentium II had obsoleted the PPro and there was no incentive to support PPro on later than 440FX.
 
I never saw the point in slot 1. So, the mfr saved on not having to implement a socket horizontally on the mobo, but on the other hand still had to provide the same socket vertically on a card. Maybe the marketing gurus thought you CPU but the mobo would live on forever. That concept went out faster than firewire.
 
To the best of my knowledge, the socket 8 slotkets only work with 440FX chipset slot 1 boards, as the later 440BX and kin did not explicitly support the PPro. It might be possible to adapt a 440BX to do so, but by the BX's time Pentium II had obsoleted the PPro and there was no incentive to support PPro on later than 440FX.

I do believe there was nothing newer was used for socket 8 boards then 440fx, But I have seen people use the slotkets on 440lx boards.
This why I asked if there was any socket 8 board with 440lx chip sets.

I never saw the point in slot 1. So, the mfr saved on not having to implement a socket horizontally on the mobo, but on the other hand still had to provide the same socket vertically on a card. Maybe the marketing gurus thought you CPU but the mobo would live on forever. That concept went out faster than firewire.
The idea was to put slower off die L2 on the card witch cut cost alot compared to full speed on die L2.
 
I never saw the point in slot 1. So, the mfr saved on not having to implement a socket horizontally on the mobo, but on the other hand still had to provide the same socket vertically on a card. Maybe the marketing gurus thought you CPU but the mobo would live on forever. That concept went out faster than firewire.

The socket 8 package was very expensive to make, and the yields became an AND operation; for a good package, the CPU die AND the one or two cache SRAM chip(s) had to pass the manufacturing QA diagnostic tests, and they couldn't be separately tested pre-assembly. So, if either failed, the whole package was a loss.

The slot 1 package costs less, and yields, now no longer an AND operation, could be higher, since the SRAM was packaged and not bare chips, and they could be separately tested from the CPU prior to package assembly. It was all economics.

But once the cache could be on-die as in socket 370, and especially since the organic PGA was developed, the driving need for the multi-chip packaging went away, and socket 370 became less expensive to produce.
 
The socket 8 package was very expensive to make, and the yields became an AND operation; for a good package, the CPU die AND the one or two cache SRAM chip(s) had to pass the manufacturing QA diagnostic tests, and they couldn't be separately tested pre-assembly. So, if either failed, the whole package was a loss.

The slot 1 package costs less, and yields, now no longer an AND operation, could be higher, since the SRAM was packaged and not bare chips, and they could be separately tested from the CPU prior to package assembly. It was all economics.

But once the cache could be on-die as in socket 370, and especially since the organic PGA was develhe driving need for the multi-chip packaging went away, and socket 370 became less expensive to produce.

I buy everything your saying. The fed unit that I worked during that period went with Dell slot 1's and they were replace hardly a year or so later. (I'm thinking early 2000's) Prior to that, there was a run of bad mobo's from Dell and they sent techs out into the field who replaced them on the spot. When I finally bailed out in 2007, Dell was still the PC of choice for a lot of the feds but the laptops depended on your specific job needs. We had a lot of HP Tough Books but that's another story.
 
Back
Top