Which revision code on the 8881?
Also, I believe the BIOSes for the 8881 have to be explicitly told that there is EDO RAM installed; there is no detection feature like on Pentium boards.
You might look over mkarcher's thread: 486 board with UMC 8881E/8886B: The winner is: EDO without L2 (if your only other option is L2 at 3-2-2-2)
Based on my understanding of the benchmarks, you only see a benefit of EDO at 60 MHz, not 40.

jakethompson1 wrote on 2023-07-02, 15:41:

Which revision code on the 8881? Also, I believe the BIOSes for the 8881 have to be explicitly told that there is EDO RAM instal […]
Show full quote

Which revision code on the 8881?
Also, I believe the BIOSes for the 8881 have to be explicitly told that there is EDO RAM installed; there is no detection feature like on Pentium boards.
You might look over mkarcher's thread: 486 board with UMC 8881E/8886B: The winner is: EDO without L2 (if your only other option is L2 at 3-2-2-2)
Based on my understanding of the benchmarks, you only see a benefit of EDO at 60 MHz, not 40.

The chip is labeled UM8881F with a date code of 9546-ETS. The BIOS is dated 12/19/95 and has options called EDO DRAM install option and EDO DRAM read speed. Trying different EDO DRAM read speed options did not help.

jakethompson1 wrote on 2023-07-02, 15:41:

...
You might look over mkarcher's thread: 486 board with UMC 8881E/8886B: The winner is: EDO without L2 (if your only other option is L2 at 3-2-2-2)
Based on my understanding of the benchmarks, you only see a benefit of EDO at 60 MHz, not 40.

Yes, that's right.
486 chipsets (or particular revisions of them) seldom take any benefit out of EDO DRAMs despite they work with EDO DRAMs at all!

In his post, mkarcher has found out that in his Biostar board there were benefits with EDO DRAMs in only 2 modes:
Cyrix 5x86 120 MHz, 60 MHz FSB with no L2 cache chips installed and up to 3x 32 MB 60 ns EDO modules and 4-2-2-2 timing (absolute winner)
Cyrix 5x86 120 MHz, 40 MHz FSB with no L2 cache chips installed and one 32 MB 50 ns EDO module and special 3-1-1-1 turbo timing

Please note which CPU settings Cyrix 5x86 he had enabled.

Disruptor wrote on 2023-07-02, 16:31:

Yes, that's right. 486 chipsets (or particular revisions of them) seldom take any benefit out of EDO DRAMs despite they work wit […]
Show full quote

jakethompson1 wrote on 2023-07-02, 15:41:

...
You might look over mkarcher's thread: 486 board with UMC 8881E/8886B: The winner is: EDO without L2 (if your only other option is L2 at 3-2-2-2)
Based on my understanding of the benchmarks, you only see a benefit of EDO at 60 MHz, not 40.

Yes, that's right.
486 chipsets (or particular revisions of them) seldom take any benefit out of EDO DRAMs despite they work with EDO DRAMs at all!

In his post, mkarcher has found out that in his Biostar board there were benefits with EDO DRAMs in only 2 modes:
Cyrix 5x86 120 MHz, 60 MHz FSB with no L2 cache chips installed and up to 3x 32 MB 60 ns EDO modules and 4-2-2-2 timing (absolute winner)
Cyrix 5x86 120 MHz, 40 MHz FSB with no L2 cache chips installed and one 32 MB 50 ns EDO module and special 3-1-1-1 turbo timing

Please note which CPU settings Cyrix 5x86 he had enabled.

Thanks, this makes feel better about using FPM memory knowing there is only a benefit in a few configurations.

jasa1063 wrote on 2023-07-02, 16:45:

Thanks, this makes feel better about using FPM memory knowing there is only a benefit in a few configurations.

Yes, you should make more benefit when you try to enable some Cyrix 5x86 features (like branch prediction) and get a stable configuration out of it.
What's the printing on the back of your CPU?

x

Disruptor wrote on 2023-07-02, 16:55:

jasa1063 wrote on 2023-07-02, 16:45:

Thanks, this makes feel better about using FPM memory knowing there is only a benefit in a few configurations.

Yes, you should make more benefit when you try to enable some Cyrix 5x86 features (like branch prediction) and get a stable configuration out of it.
What's the printing on the back of your CPU?

I already have the 5x86 configured using Peter Moss' 5x86 utility. Attached are front and back images of the CPU.

I have TD-4IP-UMC-AIO board that looks the same and it has some problems with EDO memory too. FPM works great with no speed reduction vs EDO.

DundyTheCroc wrote on 2023-07-02, 21:09:

I have TD-4IP-UMC-AIO board that looks the same and it has some problems with EDO memory too. FPM works great with no speed reduction vs EDO.

Thanks for confirmation on the EDO issue. I should have posted sooner to save myself the headache of trying to make EDO work, but it all worked out in the end:)