E10k StarFire spotted on local auction site

Yeah, I was pretty heavily involved with E10K before it was E10K ... Starfire, under NDA, already installed and running on-site, had one of the earliest, S/N was well below 100, I think it was sixty-something or so. Also did lots of acceptance/"torture" testing on such. Founds issues along the way, but Sun was always, "Oh, that's not a hardware issue, that's a software issue ... we'll patch that for you" ... and they did. Along the way, I also managed to break Solaris in a way nobody had ever broken Solaris before. ;-) Yeah, I tested very heavily and thoroughly (for months!), and, at least allegedly, we never found a *hardware* bug. Those things were beasts at the time, nobody else had \*nix at that time that could scale to that size as a single system image (and of course it could be hardware partitioned into multiple smaller systems). I think we started with relatively minimal configuration - enough to handle 2 or 3 partitions, and had about 2 or 3 cabinets/racks of storage attached to it. By the time we outgrew it, I think we had the E10K itself pretty much maxed out, or near to it, and had somewhere in the range of 10 to 15 cabinets/racks of storage direct attached to it - lasted a fair number of years (probably half dozen or more) before our largest application grew that large ... by then I think we had around half a dozen E10Ks on-site in our various data centers - so the hardware didn't go to waste at all - though one app did finally outgrow it - it was a huge "data warehouse" project, ... though it was called "data mart", for mostly political purposes (most notably because earlier someone did a data warehouse project with millions of dollars, years, on other hardware, and miserably failed, so, "data warehouse" kind'a left a bad taste and aversion in everyone's mouth, so .. it was called "data mart". And, damn smart manager of that data mart. Rather than try and such in everything and build some huge thing - at least initially, he started with very scalable platform (E10K, etc.), found the need and budget, built that highly successfully, made folks aware of that, and, lather, rinse, repeat - keep finding needs and budget and successfully adding/growing/building each; also had enough in budget to generally stay moderately ahead of needs and further budget - so things could generally be implemented quite quickly, as hardware capacity was generally already there ... and the increased budget for additional applications/needs generally built out more hardware to cover for the next needs that would come up - so was highly responsive in meeting needs/requirements with relatively short lead times, as generally the needed hardware and capacity was already there.)