I believe that even Suns of the SPARCstation-5 period used the ATX standard as far as the power connector went. The form factor of the power supply would not fit any PC of the period, but you could set an ATX supply beside the computer box with the lid open and connect an ATX power supply to the system board to run things in an emergency. (I've never actually tested this, so I may be wrong here.)
Much later the same connector is used for the voltage sense feedback in the Sun Blade 2000 -- and probably the same pinout too -- but there is a much larger connector with really serious sized wires to actually *power* the board.
BTW If you ever get a chance to open up a power supply for a Sun Blade 1000 or 2000, do so. It has to be the most beautifully constructed power supply of any that I have ever seen.
[ ... ]But -- before the ATX came along -- with the IBM AT computers, there was one thing which was not done right in the power supply. The other voltages (12V and others) were dependent on the 5V being loaded to a certain percentage of max before you got regulation on the others. As a result, if you had an AT with only one hard drive, you needed to run a big load resistor on the +5V pins of the second drive's power connector (and the resistor was mounted where the drive would mount, so when you added the second drive, the resistor would be removed.)
Sounds like a good design -- which may be why Sun started using it in the SS-5 and similar machines.
Well ... not *any* power supply. But well designed ones, including lab bench supplies, yes. The nicer ones even offered adjustable current limit, so it could be used as a constant current supply within the voltage swing range of the supply.
Enjoy, DoN.