I used to have one. If you boot your PC with two graphics cards, one whatever that posts and the other is this one, you'll see that the device manager recognizes the Sun Permedia 2 too and drivers can be installed. That's as far as I got with it, then ran out of patience and sold it.

Edit. Allegedly if you do it like this, than it should post after the drivers load or something, but it didn't for me.

RandomStranger wrote on 2024-01-31, 12:01:

I used to have one. If you boot your PC with two graphics cards, one whatever that posts and the other is this one, you'll see that the device manager recognizes the Sun Permedia 2 too and drivers can be installed. That's as far as I got with it, then ran out of patience and sold it.

So the problem would be the firmware?

That was my suspicion too, but never attempted to tamper with the firmware/BIOS.

George Razvan wrote on 2024-01-31, 12:03:

RandomStranger wrote on 2024-01-31, 12:01:

I used to have one. If you boot your PC with two graphics cards, one whatever that posts and the other is this one, you'll see that the device manager recognizes the Sun Permedia 2 too and drivers can be installed. That's as far as I got with it, then ran out of patience and sold it.

So the problem would be the firmware?

The firmware is fcode for Openboot on a SPARC system but should also work on a PPC Mac and some IBM Power systems. It is not an x86 BIOS.

I spotted one of these cards very cheaply, so I picked it up to see what I could do with it.

Happy to say that with the help of UniFlash, I was able to flash a IBM compatible BIOS to the card and it works! As the original BIOS was 64K, it was simply a matter of copying the IBM 32K ROM on top of itself to produce a valid 64K image. And thankfully UniFlash was able to talk to the EEPROM on the card.

To flash the card:

• Pop the card in alongside an existing VGA card.
• Run uniflash from DOS with the -pcirom switch.
• Select the card from the list of devices (you're looking for the device with a vendor ID of 3D3D and a device ID of 0009).
• Choose Flash BIOS image INCLUDING bootblock and enter the name of the ROM file.
• Once the flash has completed, remove your existing VGA card. You should now have video output on the Raptor card. Note that somewhat counterintuitively, you should have JP2 configured as VGA DISABLED. I don't think the other jumper settings matter however.

I also include the original SPARC BIOS that I took a copy of from the card before tinkering with it.

Hi there! I happen have one of these cards in my pile of old Sun parts for at least the past 20 years. I happen to see your post from last year about flashing one of these to use in a PC, which would be fantastic since I have a P233MMX system that needs a decent PCI video card. I followed your instructions and did everything the same but even though it says in the card selection screen that is has a 64k bios, when I select the card it says "1024 Pages of 128 Bytes (128k)" and it WON'T let me flash your bios file! What is going on here? I can't figure out why since its the same card and the Uniflash utility is telling me that its 64k, its saying its 128k. Thank you for ANY help here, there are a ton of Uniflash options as well as different versions and I'm not sure if I should be trying anything else.

The 128K card is an Adaptec SCSI card by the looks of it - I'd try removing any other cards that you can for the flashing process.

No, I'm not selecting that one. I'm selecting the 3D3D vendor card, which is then saying its a 128k BIOS file it needs. I pulled it out of my P233 and put it into my P100 computer which has no other PCI cards and its the exact same thing. The selection screen shows the 3D3D card as 64k, but selecting it says its 128k. Would it be possible for you to take that exact Bios file and do whatever you did to duplicate it from 32k so it equals the full 128k? I tried a few things but it was never able to write the file successfully.

Here is the backup of the BIOS i took from the card before trying anything, maybe it will be useful to see whats going on here.

Ahh sorry, my mistake! I can see the chip on the board is definitely 128k (AT29C010A).

You can just run copy /b p2v_IBM.bin+p2v_IBM.bin p2v_128k.bin to copy the file on top of itself to create a 128k file, but I also attach a 128k version here for convenience.

Unfortunately it seemed to start to flash but then had some red blocks towards the end and ended with this error.

That seems about half way. TBH, I’d give it a try. If you’re feeling adventurous you can replace the flash chip with some soldering. I’d try reflashing a few times too just to see if the chip improves.

Sadly it’s not unusual for chips to fail, but it is fixable.

I wonder if the chip is configured to only work in a 64k mode or something?

Also, not sure if you have any shadowing enabled in the bios that might need to be disabled?

I'll check all of that, thanks again for your help. I don't really have any soldering skills so I'll probably just flash it back to its original Sun Bios and stick it back in the parts pile in case I get a SunFire or something like that again.

Video bios shadowing was enabled but turning it off didn't make a difference. The card also does not work (just beeping and no video), so whatever that copying trick did to make it work for the 64k bios I suspect they probably had 128k chips on hand and used them for some cards and somehow binned them down to 64k. It's confusing Uniflash which is trying to use the whole thing (which is why its being reported and detected at two different sizes most likely).

Sorry this isn’t working 🙁 Hopefully there’s a way to do it. Can you force a different chip size in Uniflash?

This bios and flashing method also worked on my Melco/Buffalo WHP-PS8S Permedia2v PCI. Dip switch 1 had to be OFF (up toward edge of card) for its bios to be writeable by uniflash.