Attendees

1. @Thomas Watteyne

2. @David Burnett

3. @Titan Yuan

4. @Fil Maksimovic

5. @Sara Faour

6. @Kristofer PISTER

Action Items

Updates

• workshop @Thomas Watteyne

  • thanks @David Burnett for filling in page

  • thanks @Fil Maksimovic for agreeing to write intro paragraph to CFP

  • we have a website! https://project.inria.fr/crystalfree/

  • rooms are booked for the 3 days Sep 12, 2023 - Sep 14, 2023

  • can we rename to “Invited Workshop on Crystal-Free Radios, Chip-Scale Systems and Swarms”?

• @David Burnett

  • 5-bond SCuM board w/ 337 battery holder sent for mfg

    • Student (Luis) made an argument for GSG bonds instead of just S for antenna so the Poynting vector is in the direction of the wirebonds

  • Can likely get away with much cheaper ENIG surface finish

  • Found room temperature stored, room temperature cured silver epoxy w/ 10 min, 20 min or 4 hour working time https://www.mouser.com/ProductDetail/MG-Chemicals/8331D-14G?qs=doiCPypUmgEsDWzvEstgPw%3D%3D

  • Student (Jacob) has been extensively testing using only a single supply, will add tests at 1.4 V and 1.3 V soon and test from a 337 battery before sending out chips for wirebonding

  • Jacob also assembled a SCuM “kit” DB is taking to APL-UW in Seattle tomorrow

• @Sara Faour / @Fil Maksimovic

  • attempt at using code to generate either a unique identifier, ideally PUF

  • extracting today 64 bits

    • code actually reads 8 times 8 bytes (64B), but no technical difficulty to read much more

  • sweep 900mV to 1.1V, same value

  • need to test against temperature

  • @Kristofer PISTER : ramp rate on the VDD voltage influences PUF initial state

  • PUF vs. indentifier: hamming distance and entropy, testing code

  • quite a bit of SOTA

  • code: https://github.com/Sara-Fa/scum-test-code/tree/develop/scm_v3c/applications/sram_read

  • Considering BLAKE2s (BLAKE optimized for 32 bit CPUs) hashing algorithm https://www.blake2.net/

    • BLAKE3 faster but designed for x86 architecture