nforce4max wrote:

Yes it looks to be a decent board, apples to apples or apples to oranges but at the end of the day the gains from adding the real cache are small potatoes. Just be happy that you got a socket 3 board with pci and just keep looking around for other boards.

I don't actually own one, but I'm considering getting one partly for its infamous reputation. 🤣 I'm intending to put in a machine that I'll use as a conversation piece more than anything.

soviet conscript wrote:

I havnt had any issues with it. I read some people had stability issues including EMS errors and system lockups but they may be with older revisions.

My Rev 3.4 gets a freeze from low ram timings when trying to do a soft reset

Speedsys score of 59.79 ain't bad though 😀

I have one of those fake cache 486 PCI/VLB boards. You can see there are no traces to the soldered on cache chips. From what I recall if you did install one of the special (rare specific to this board) cache coast modules it would use it.

It's installed inside a mini AT tower and I am probably going to sell it one day.

does anyone know if fake COAST modules are common or am I missing something? I finally got a hold of a 256kb cache module for the M919 for a reasonable price. I installed the module in the board today and it defiantly "sees" it because on boot instead of saying "write back cache" (yhea right) is now says "256kb cache" BUT when I run cachechk it still isnt seeing anything and only reports the 16kb L1 cache only.

Perhaps its possible to update the bios with a bios from a M919 with working cache?

I had a M919 and it would freeze with certain EMM386 switches (I suppose because the BIOS lies about certain functions being available?), but otherwise seemed stable enough.