The EEPROM board uses pads on its PCB to create an "edge connector" with 8 pads. Apparently only 4 pads are needed, and presumably there are twice as many for r

The EEPROM board uses pads on its PCB to create an "edge connector" with 8 pads. Apparently only 4 pads are needed, and presumably there are twice as many for redundancy. Physically, the PCB happens to fit into a USB socket with close enough spacing to make good contact. This isn't surprising, because it's a very common trick to make USB plugs out of PCBs. Electrically, it's not USB though. It's two pins for power, and two for I2C clock and data. It might happen to have the same voltage and pin mapping for power as USB, though, so plugging in to a USB port on a computer might not blow anything up. A microcontroller with a USB host port might even be able to bit-bang I2C over the appropriate pins, but that would be rather clever...