1. Philips Research ApplyingAsynchronous Circuits in Contactless Smart Cards Applying Asynchronous Circuits in Contactless Smart Cards Joep Kessels, Torsten Kramer Gerrit den Besten, Ad Peeters, Volker Timm

2. Philips Research Esprit project Descale Period: 1998-1999 Participants: Philips Semiconductors, MAZ, Philips Research Goal: find out advantages of asynchronous circuits in contactless smart cards

3. Philips Research Outline Designing asynchronous circuits Contactless smart cards Applying asynchronous circuits in contactless smart cards

4. Philips Research VLSI programming of Handshake circuits Designing asynchronous circuits (handshake circuits) in a high level programming language (Tangram) using a compiler for translation (transparent)

5. Philips Research Tangram Handshake circuit Sequential composition A A ; B ; B

6. Philips Research Tangram Handshake circuit Parallel composition A || B B A ||

7. Philips Research Synchronous versus asynchronous Clock driven Demand driven Less average power Central clock Distributed Smaller current peaks handshakes Less EM emission Clock timed Self timed Performance adaptation to supply voltage (1..3 V)

8. Philips Research Contactless smart card Tuned circuit: –Power –Clock –Communication

9. Philips Research Mifare (ISO standard) - Proximity card (10 cm) with two way communication - Power: few mW; Transaction time: 200 msec - 70 M cards sold - Clock: 13.56 MHz; Bitrate: 106 Kbit/sec

10. Philips Research Differences in power characteristics Minimizing Average Peak Power Supply Constant Fluctuating Voltage

11. Philips Research Digital Circuit Peripherals: –DES –RSA –UART Memories: –2 kbyte RAM (10 nsec) –32 kbyte EEPROM (read/write 180/4000 nsec) –38 kbyte ROM (30 nsec)

12. Philips Research Power 80C51 in time domain SynchronousAsynchronous

13. Philips Research Power 80C51 in frequency domain Synchronous Asynchronous 0 100 200 300 400 MHz 0 100 200 300 400 MHz

14. Philips Research Performance adaptation asynchronous 80C51

15. Philips Research Descale chip 5-layer metal 0.35  m 18 mm 2

16. Philips Research Area/Power contactless digital circuitry Async about 12% of contactless digital circuit area

17. Philips Research Effect asynchronous design

18. Philips Research Power regulator

19. Philips Research Improvements in Tangram Toolset Redefinition Tangram –communication through variables Use of conventional tools for data-path part –Optimizer & Technology mapper to reduce area (10%) –Timing analysis tool to tune matching delays (up to 50%)

20. Philips Research Conventional solution Synchronous digital circuit with fixed speed –superfluous power thrown away –too little power: transaction is canceled Performance 80C51 limited by power received Buffer capacitor of several nF (large area)

21. Philips Research Advantages asynchronous design Maximum performance for power received –power efficiency: factor 2 –adaptation property: factor 2 More robustness and/or smaller buffer capacitor –smaller current peaks –adaptation property

22. Philips Research Conclusion Results so convincing that a product is being designed based on these asynchronous circuits

23. Philips Research Mifare Applications Seoul: six million bus cards Lufthansa: Frequent Flyers cards China: highway toll cards Brasil: cards for civil servants (identification & electronic purse) Shell: Mifare technology in car keys

24. Philips Research Modifications in 80C51 Instruction prefetching (30% more performance) Early write completion Immediate halt signal Quasi synchronous mode (performance 50% of free-running mode)

25. Philips Research DES convertor Transaction contains up to 10 DES conversions Software conversion : 10 msec, 30  J Hardware conversion: 1.25  s, 12 nJ Area 3,250 GE - 57% combinational logic - 35% latches/flipflops - 8% delay matching and C-elements

26. Philips Research Power 80C51 and DES co-processor @ 3.3V