Intuition wrote:Wait, does the Octane lyrics imply you did find a way to plug it into an Octane card cage?
Unfortunately not, but they didn't write a song about the O2... I still welcome anyone who is interested to give it a go in another SGI, again, somebody out there might have a better understanding of the drivers than I do, and a D1 in a Tezro would be a truly capable SGI, both in IRIX software and modern applications. I did not write the drivers for the D1, again, it exposes itself as a network interface and is based on a reverse-engineered SGI device, so the drivers are already there.
Intuition wrote:Seriously though, I am pretty excited about this modern o2 biz.
So am I! The true breakthrough for "modern SGIs" would be if we could someday replicate the Indy chipset on an FPGA and find a supply of R4000s, but I'm not holding my breath for that one (nor do I have enough of an understanding of FPGAs to do it myself). The IP22 documentation is out there! Amusingly, the only person here I seem to remember ever referencing FPGAs was everyone's favorite Italian redditor, foretold in song and prophecy, so I'm not holding my breath for THAT.
I talked this over with Ian Mapleson, and I think our best chance at replicating SGIs right now would be to build new Indigo2s from all the cheap motherboards that seem to be floating around. No part of that system is expensive, except the case (which is almost always surrounding a complete Indigo2 inside). It would not be particularly sustainable, but I bet with a D1-level effort I could produce five or ten brand-new Indigo2s from used boards/CPUs/graphics and replica parts. One thing that helps on the IP22 boxes is their relatively standard RAM. I would probably also redesign the SCSI drives to be an SD-based emulator, use a smaller power supply, etc. End result could possibly be around the size of an Indy, if the EISA/GIO riser was redesigned, but I haven't looked into this quite so extensively. A project for another day.