CW310/Bergen Board Demo Results Question

Hello,

I am testing the provided demo project with a simple AES on the CW310/Bergen board. I noticed that I cannot break the AES with 5k traces or even 50k traces. At 50k about half of the bytes have been recovered. I believe this is essentially the same design as the one used for the CW305 board, and that breaks within 5k traces easily. I am also using the same attack model as what is used for the CE305 demo. Do you have an explanation for this behavior?

Thanks,

Jackie

  1. Exactly which bitfile are you using?
  2. Do you have the “normal” version or the “SCA” version of the CW310?
  1. I am using my own bitstream generated from the AES project in the cw310-Bergen-board repo.

  2. I have the K410T normal board, I didn’t know there were two variations. Is there documentation on the differences between the two versions?

Yes, as explained here the normal version has decoupling capacitors mounted, which makes it less suitable for side-channel work.

We can convert the board for you (remove the decoupling capacitors); contact sales@newae.com for pricing.

1 Like

Just to confirm, these are capacitors C21 through C29 and C180 on the schematic right? I assume so since the notation implies they would not be on the SCA version, but wanted to double check.

Yes, that is correct.

1 Like

Hello, I’d like to know if I can change the CW340/Luna board normal version to the SCA version using the same method.

Yes- the capacitors to switch from normal to SCA are on the CW341 daughtercard:
they are the 10 caps connected to +1V0 on page 2 here: cw340-luna-board/cw341-kintexus-ku095/NPCA-CW341-KINTEXUS-03.PDF at main · newaetech/cw340-luna-board · GitHub
(C36, 40, 35, 44, 39, 29, 42, 26, 21, 20)