@Shane Dowd: The actual pad cells in the sky130 I/O library are way over-designed for most purposes and have a massive set of power rings. Taking the general-purpose everything-but-the-kitchen-sink I/O cell as a typically-sized cell, it's 81um x 200um (so Larry's estimate is pretty accurate). What the library doesn't have is any cell that can be used to stagger I/Os. But you can put a bond pad anywhere you want as long as you want to design all the ESD circuitry around it yourself; the bond pad size is 65um x 75um. If you want a probe pad you can go much smaller than that---See the example Caravel user designs from NIST/U. Michigan which have all internal pads and were designed for probing on a probe station. In fact, if you look through the list of past MPW projects at efabless.com, you'll find a number of them that made their own internal padframe. I made my own set of simplified pad cells which was never characterized or zap-tested, but they worked. All I did was to throw together the existing bare bond pad with some ESD diodes under it, designed on basic principles, and I pulled the clamp design from the sky130 I/O cell library.