Yes. Most IDE controllers don't need their own BIOS as they just hook into standard BIOS IDE interface. You might need to disable the onboard IDE though. Now the question is if you already have an onboard IO why are you installing another one in the slot?

VL-Bus system at 50 MHz sounds like a big ask. You'll be lucky if you get it working at 40 MHz, especially with 2 VL-Bus cards.

mpe wrote on 2020-01-10, 12:33:

VL-Bus system at 50 MHz sounds like a big ask. You'll be lucky if you get it working at 40 MHz, especially with 2 VL-Bus cards.

Also, instead of overclocking a rare CPU, just use a DX2 if you want more speed. It makes absolutely no sense to use an UMC U5S-SUPER40 at all if all you are aiming for is speed. I assume you know that the U5S lacks an FPU, right?

I tried the card because it was sold as untested, so I had to test if it was actually working or not, since I also lack a proper VLB mb (which I'd like an advice about what model look for).
That being said in the card there is jumper setting which divide the clock, it may prove useful.
To answer the question why the rare cpu, pretty easy, because it's rare and it was faster clock to clock than other 486 cpu, despite lacking a fpu, as shown in this video that I mentioned early:
https://www.youtube.com/watch?v=GfO0IFMXUT8

Also the guy was able to run the cpu at 50mhz without issues, was he so lucky?

Problem is not the CPU running at 50 Mhz, but the VLB. This might work with some Video boards but will end up in data corruption pretty sure using an ide controller. VLB was designed for 33 MHz, 40 and 50 Mhz usually needed wait states, or on board peripherals.
eisapc

eisapc wrote on 2020-01-10, 13:19:

Problem is not the CPU running at 50 Mhz, but the VLB. This might work with some Video boards but will end up in data corruption pretty sure using an ide controller. VLB was designed for 33 MHz, 40 and 50 Mhz usually needed wait states, or on board peripherals.
eisapc

Is this going to happen even if I select 50mhz on the ide controller, it should divide the frequency so it will go to 25mhz, is this assumpion correct?
Picture: https://i.ebayimg.com/images/g/UuEAAOSw7gpdmdLz/s-l1600.jpg
Do you have any specific motherboard I should choose from?
I was thinking to get this one (https://drive.google.com/file/d/1KonJu_LsDJuU … iew?usp=sharing) but if you have any other model to suggest, ill wait.

Thanks.

My first PC build was a 486DX 50 VLB and it would not work with a VLB Graphics card. I had to buy a fast ISA graphics card, Tseng ET4000. I had a computer shop at the time and I am sure I would have tried all the VLB graphics cards I had.

I don't think the card can divide the VL-Bus frequency. It has to communicate at 50 MHz on the host side as this is how the CPU is driving the bus.

I think what the 33/50 MHz jumper does is to calibrate command cycle length on the IDE side of the interface. So that the PIO4 cycle length is roughly 120ns as per ATA spec, etc. The card is getting the clock from the VL bus so it useful to know roughly how long the clock is. Otherwise PIO ratings will be way out of spec.

You'll still need to set wait-states on the motherboard in order to run it reliably (if at all possible) which will slow down the system.

Looking at the video you linked my own result from my ordinary DX2-66 are quite a bit faster in benchmarks than what is in the video. Could that be due to the extra waitstates?

40Mhz should be fine with everything without any wait states. 50Mhz will have problems with IDE and some video cards.

mpe wrote on 2020-01-10, 20:17:

I don't think the card can divide the VL-Bus frequency. It has to communicate at 50 MHz on the host side as this is how the CPU […]
Show full quote

I don't think the card can divide the VL-Bus frequency. It has to communicate at 50 MHz on the host side as this is how the CPU is driving the bus.

I think what the 33/50 MHz jumper does is to calibrate command cycle length on the IDE side of the interface. So that the PIO4 cycle length is roughly 120ns as per ATA spec, etc. The card is getting the clock from the VL bus so it useful to know roughly how long the clock is. Otherwise PIO ratings will be way out of spec.

You'll still need to set wait-states on the motherboard in order to run it reliably (if at all possible) which will slow down the system.

Looking at the video you linked my own result from my ordinary DX2-66 are quite a bit faster in benchmarks than what is in the video. Could that be due to the extra waitstates?

Thank you for the clarification, so what I need is basically a mb that let me select the wait states between 0 and 1.
I do not know about benchmark results, but you could be definitely right

I know where you can get a ARK Logic ARK1000VL VLB probably for pretty cheap, I am not the seller. But I know someone who has like 100 of them.

like how much is cheap?

I received the ark logic card, let's say i'm... disappointed about cheap... shipping price if we can call so...

Back in the 90s, I ran a Diamond Stealth 64 DRAM VLB 2MB (S3 Trio64) with a 486DX-50. I'm not sure if this will be true of every example of this card, however.

You may get it to work @ 50MHz, but it's a big ask. The problem is not the cards per-se, it's the bus. VLB was a hack. I love my VLB DX2 build, don't get me wrong, but it was some cowboy engineering for sure. I had to try two different motherboards and some slot reordering to get a VLB graphics card, I/O card, and SCSI card to work together.... at 33MHz.