Pretty sure all revisions only support FPM. I know up to B4 is said to be FPM only.

I have a Lucky Star LS486e Rev D and it runs EDO very nicely thanks. Chipset revisions are 496 PR 497 NU

https://www.elhvb.com/webhq/models/486pci/ls486ed.htm

debs3759 wrote on 2021-03-04, 11:55:

Pretty sure all revisions only support FPM. I know up to B4 is said to be FPM only.

I don't know where this belief comes from...

See:

1486 Green PC VESA/ISA/PCI Chipset
2SiS 85C496/497
3Preliminary
4Rev 3.0
5July 1995

gnif wrote on 2021-03-06, 07:13:

I don't know where this belief comes from...

The original datasheet is in colors. This line about EDO DRAM is in purple. The purple colored sections are only implemented in later chipset revisions.

Sorry for necropost.

pc2005 wrote on 2025-07-13, 05:03:

gnif wrote on 2021-03-06, 07:13:

I don't know where this belief comes from...

The original datasheet is in colors. This line about EDO DRAM is in purple. The purple colored sections are only implemented in later chipset revisions.

Sorry for necropost.

while its true that sis496 and some other pci 486 chipsets claimed support for edo dram, i doubt if the boards actually got edo mode running properly, and those that managed to power on are likely just running in fast page mode with identical performance.
i heard that ALL 72pin edo simms for PC are actually defective by standard(but not the 168pin edo dimms for servers), lacking a few signal pins required to get edo running properly. the solution for this on 586 boards is buffer chips near simm slots(or integrated in 430hx and later northbridges), but i doubt if any 486 boards took any effort on that.

noshutdown wrote on Today, 02:00:

pc2005 wrote on 2025-07-13, 05:03:

gnif wrote on 2021-03-06, 07:13:

I don't know where this belief comes from...

The original datasheet is in colors. This line about EDO DRAM is in purple. The purple colored sections are only implemented in later chipset revisions.

Sorry for necropost.

while its true that sis496 and some other pci 486 chipsets claimed support for edo dram, i doubt if the boards actually got edo mode running properly, and those that managed to power on are likely just running in fast page mode with identical performance.
i heard that ALL 72pin edo simms for PC are actually defective by standard(but not the 168pin edo dimms for servers), lacking a few signal pins required to get edo running properly. the solution for this on 586 boards is buffer chips near simm slots(or integrated in 430hx and later northbridges), but i doubt if any 486 boards took any effort on that.

We were just talking about this recently: Re: I see no performance difference between EDO and FPM

The idea that these 486 chipsets don't actually support EDO, or only run EDO in a degraded FPM timing mode keeps popping up.
It could be that the BIOS is overly conservative, and the fastest EDO timing is either a hidden option, hardcoded into a chipset register by the manufacturer using Modbin or AMIBCP, or perhaps determined by code to measure the bus speed, but in whatever way, inaccessible to you as the user.
Or, perhaps the SiS 496 is the culprit. Looking at its datasheet, at 25 MHz (should correspond to "fastest") FPM and EDO are both 3-2-2-2 burst read hit. At 33 MHz, FPM is 4-3-3-3 and EDO is 4-2-2-2. Unimpressive considering that Intel 430HX runs EDO at 5-2-2-2 at 66 MHz.
The ALi FINALI chipset, known for impressive EDO RAM performance for a 486 chipset, suggests X-1-1-1 burst timings for EDO, and X-2-2-2 for FPM.
The linked thread discusses the UMC 8881 having a 3-1-1-1 EDO mode.