I've been a part of the IBM team that designs those HSMs since long before the 4758 - in fact, I was part of the original ad-tech group that did the research leading to our first HSM with that kind of packaging, the 4755 which was announced in 1989. While there are definitely a lot of similarities over generations, there has been an evolution and the current IBM HSMs are quite different from the 4758. The technology used in the IBM 4769 HSM is covered to some degree in the paper "The IBM 4769 Cryptographic Coprocessor" In the September/November 2020 issue of the IBM Journal of Research and Development.. (https://ieeexplore.ieee.org/document/9138701) I'd be happy to forward the PDF file to anyone who is unable to get to the web site.
From: [email protected] To: [email protected] Date: 11/06/2020 02:57 AM Subject: [EXTERNAL] Tech Digest, Vol 70, Issue 1 Sent by: "Tech" <[email protected]> Send Tech mailing list submissions to [email protected] To subscribe or unsubscribe via the World Wide Web, visit https://lists.cryptech.is/listinfo/tech or, via email, send a message with subject or body 'help' to [email protected] You can reach the person managing the list at [email protected] When replying, please edit your Subject line so it is more specific than "Re: Contents of Tech digest..." Today's Topics: 1. Disassembly, and production of HSMs (Joachim Str?mbergson) 2. Re: Disassembly, and production of HSMs (Peter Gutmann) ---------------------------------------------------------------------- Message: 1 Date: Fri, 6 Nov 2020 07:48:28 +0100 From: Joachim Str?mbergson <[email protected]> To: "[email protected]" <[email protected]> Subject: [Cryptech Tech] Disassembly, and production of HSMs Message-ID: <[email protected]> Content-Type: text/plain; charset="utf-8" Aloha! (Sharing some HW pr0n this Friday.) Was pointed to a very interesting paper that describes how IBM manufactures FIPS Level 4, PCIe based HSMs. Quite a lot of resin and glue required. https://www.co.tt/files/tamper_proof_encryption_smta.pdf The paper nicely complements an older set of images that shows somebody disassembling an older version of the IBM HSMs: https://www.flickr.com/photos/flyingspores/albums/72157628094589706/with/6331000908/ -- Med v?nlig h?lsning, Yours Joachim Str?mbergson ======================================================================== Assured AB ======================================================================== -------------- next part -------------- A non-text attachment was scrubbed... Name: OpenPGP_0x5DDC7C542422EC8D.asc Type: application/pgp-keys Size: 20513 bytes Desc: not available URL: < https://lists.cryptech.is/archives/tech/attachments/20201106/f7942dc2/attachment-0001.key > -------------- next part -------------- A non-text attachment was scrubbed... Name: OpenPGP_signature Type: application/pgp-signature Size: 840 bytes Desc: OpenPGP digital signature URL: < https://lists.cryptech.is/archives/tech/attachments/20201106/f7942dc2/attachment-0001.sig > ------------------------------ Message: 2 Date: Fri, 6 Nov 2020 07:56:39 +0000 From: Peter Gutmann <[email protected]> To: Joachim Str?mbergson <[email protected]>, "[email protected]" <[email protected]> Subject: Re: [Cryptech Tech] Disassembly, and production of HSMs Message-ID: <[email protected]> Content-Type: text/plain; charset=WINDOWS-1252 Joachim Str?mbergson <[email protected]> writes: >Was pointed to a very interesting paper that describes how IBM manufactures >FIPS Level 4, PCIe based HSMs. Quite a lot of resin and glue required. Interesting that this paper seems to completely erase from history the IBM 4758, which was more or less the same thing done twenty years earlier. I wonder what the politics behind that was? Peter. ------------------------------ Subject: Digest Footer _______________________________________________ Tech mailing list [email protected] https://lists.cryptech.is/listinfo/tech ------------------------------ End of Tech Digest, Vol 70, Issue 1 ***********************************
_______________________________________________ Tech mailing list [email protected] https://lists.cryptech.is/listinfo/tech
