> I miss most from when I was at Sun was the SB2000 (2x12.Ghz, 8GB, 2x > 73GB, 2x XVR-1200 - either that or XVR-1000.) > > I've searched EBay, but they're not cheap I should have pulled the > trigger on one of a few I saw for about $900, but I was still laid-off > then and couldn't justify the expense. Now I wonder if the SB2500 > wouldn't be better machine. > > -Kyle > > %m
Only a quick response in a hurry: Yes, I think the Blade 1000/2000 is the best workstation Sun ever built. The SB2500 certainly not ... Plus: The Blade 2000 uses real UltraSPARCIII (open up the module and you will find that its die is pin-compatible to the USIV, that's the REAL thing). The SB2500 uses the USIIIi, which is bad. Look up the old SPEC benchmarks: The fastest SunBlade2000 is faster than any SB2500_red. Plus: The SB2500's are still above $1K on ebay (without much mem), while you get a cheap SB1500 starting at $150. And later you can plug more mem into it. XVR-600?? Nope, Xorg has no drivers for all the 3Dlabs Wildcat chipsets, due to the restrictiveness of 3Dlabs. On x86 Xorg can use vesa_drv to operate on those cards (even, for example on Linux_x86 you don't have native drivers). On SPARC we are still lost. That's the reason why we cannot run SPARC-Xorg on any of the following: Extert3D, Expert3D Lite, XVR-500, XVR-600, XVR-1200. That's really a pity. I must give vesa_drv a new try. We must find out why exactly vesa_drv does not work on any SPARC- or PowerPC os. It probably depends on some x86-Bios interrupts, but on which one? Why doesn't xf86emu emulate them? Some general fallback-driver like vesa_drv on x86 would be too nice to have on a SPARC, too. %martin
