What is a smart card?

Julien SIMON

Cards come in two flavors: memory cards and processor cards. Memory cards have no CPU and thus have no processing power: they are just used to store data. Processor cards have a CPU, which enables them to run on-board applications: that's why they are called "smart cards".

Most cards need to be inserted into a card reader (also called card acceptance device, or CAD). However, some cards are contact less and simply need to be close enough from a reader for a given period of time.

For contact cards, the visible golden part is not the chip: these are the electrical contacts, through which the card is powered and can talk to the outside world. If you want to see the chip, you have to remove the contacts (THIS WILL DESTROY YOUR CARD!). The chip holds the following items:

  • The CPU. It usually is an 8-bit processor, running at an internal clock speed of 5MHz.
  • An optional cryptographic coprocessor, to speed up crypto operations.
  • ROM (Read-Only Memory): it can't be altered and holds the runtime environment as well as default applications. Typical amount is 64Kb.
  • EEPROM (Electrically Erasable Programmable Read-Only Memory): it's persistent and holds applications loaded after the card has been issued, as well as application data. Typical amount is 32Kb.
  • RAM (Random Access Memory): it's volatile and holds application stack, as well as transient data. Typical amount is 2Kb.
  • I/O lines, to enable the card to talk to the reader. This is typically a set of serial lines.

Future cards are likely to use a 32-bit RISC processor, Flash Memory or FeRAM instead of EEPROM, and USB instead of serial lines.