Post by thread:[EE] Cryptographic ICs

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi guys, Do any of you know of ICs one can buy to perform cryptographic primitives (block ciphers, hashes, etc.)? I looked around Digikey in the categories that seemed reasonable, but couldn't seem to find any such devices. I know I've heard of one before (even seem its datasheet), but I can't seem to find the part now. Thanks! Chris -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.17 (GNU/Linux) iEYEARECAAYFAk34TXgACgkQXUF6hOTGP7eNzQCeNHhB7bQJskpFO0bSYFHfRLwc aVEAoIw8dG2VUBz5DTENUeGX3HY1YtWm =IZqG -----END PGP SIGNATURE-----

> Hi guys, > Do any of you know of ICs one can buy to perform cryptographic > primitives (block ciphers, hashes, etc.)? I looked around Digikey in > the categories that seemed reasonable, but couldn't seem to find any > such devices. I know I've heard of one before (even seem its > datasheet), but I can't seem to find the part now. You may be thinking about the 8048 chip that Intel programmed to do DES encoding/decoding. I believe it was real slow (equivalent to around 100 baud IIRC) and required all sorts of clearances to be allowed to purchase it. These days I think everyone 'just' puts the relevant crypto engine inside an FPGA - its fast, relatively secure and makes for one chip with everything including buffer ram in it. There is code around for doing it on micros, Microchip sells a CD or download to go with their various app notes of the crypto schemes they support. The CD is a nominal cost, but they have to do it that way so that distribution of the source code can be controlled for ITAR purposes. -- Scanned by iCritical.

AVR has a DES instruction, so it should not be too hard to implement a Triple DES which performs reasonable fast. >From the DataSheet: The module is an instruction set extension to the AVR CPU, performing DES iterations. The 64-bit data block (plaintext or ciphertext) is placed in the CPU register file, registers R0-R7, where LSB of data is placed in LSB of R0 and MSB of data is placed in MSB of R7. The full 64-bit key (including parity bits) is placed in registers R8-R15, organized in the register file with LSB of key in LSB of R8 and MSB of key in MSB of R15. Executing one DES instruction performs one round in the DES algorithm. Sixteen rounds must be executed in increasing order to form the correct DES ciphertext or plaintext. Intermediate results are stored in the register file (R0-R15) after each DES instruction. The instruction's operand (K) determines which round is executed, and the half carry flag (H) determines whether encryption or decryption is performed. However, I agree with Alan that an FPGA would be a better choice for a better ciphering and/or hashing algorithm. Tamas On Wed, Jun 15, 2011 at 9:26 AM, <spam_OUTalan.b.pearceTakeThisOuTstfc.ac.uk> wrote: {Quote hidden}> > Hi guys, > > Do any of you know of ICs one can buy to perform cryptographic > > primitives (block ciphers, hashes, etc.)? I looked around Digikey in > > the categories that seemed reasonable, but couldn't seem to find any > > such devices. I know I've heard of one before (even seem its > > datasheet), but I can't seem to find the part now. > > You may be thinking about the 8048 chip that Intel programmed to do DES > encoding/decoding. I believe it was real slow (equivalent to around 100 baud > IIRC) and required all sorts of clearances to be allowed to purchase it. > > These days I think everyone 'just' puts the relevant crypto engine inside > an FPGA - its fast, relatively secure and makes for one chip with everything > including buffer ram in it. > > There is code around for doing it on micros, Microchip sells a CD or > download to go with their various app notes of the crypto schemes they > support. The CD is a nominal cost, but they have to do it that way so that > distribution of the source code can be controlled for ITAR purposes. > -- > Scanned by iCritical. > >

Simply use a PIC24, dsPIC or PIC32 and program it with the freely available algorithms. Here in Brazil we cannot buy anything that includes hardware cryptography, so we use standard MCUs and put the cryptography in the firmware. A dsPIC running at 40MIPS can do a great deal of AES encryption/decryption per second. Isaac Em 15/6/2011 03:13, Christopher Head escreveu: > Hi guys, > Do any of you know of ICs one can buy to perform cryptographic > primitives (block ciphers, hashes, etc.)? I looked around Digikey in > the categories that seemed reasonable, but couldn't seem to find any > such devices. I know I've heard of one before (even seem its > datasheet), but I can't seem to find the part now. > > Thanks! > Chris

On 15/06/2011 07:13, Christopher Head wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi guys, > Do any of you know of ICs one can buy to perform cryptographic > primitives (block ciphers, hashes, etc.)? I looked around Digikey in > the categories that seemed reasonable, but couldn't seem to find any > such devices. I know I've heard of one before (even seem its > datasheet), but I can't seem to find the part now. > many MIPs and ARM SoC used for Router/Wifi boxes have HW crypto accelerators. Needed to handle WiFi at reasonable speed. Start with looking at Broadcom and Texas chips/cpus for Router/WiFi

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Tue, 14 Jun 2011 23:13:09 -0700 Christopher Head <.....headchKILLspam@spam@gmail.com> wrote: > Do any of you know of ICs one can buy to perform cryptographic > primitives (block ciphers, hashes, etc.)? I looked around Digikey in > the categories that seemed reasonable, but couldn't seem to find any > such devices. I know I've heard of one before (even seem its > datasheet), but I can't seem to find the part now. Thanks for the pointers everyone; it seems you haven't seen the kind of thing I was looking for. Someone showed me a chip that would just accept data on an input and produce a hash of the data on the output, but I can't figure out who showed me the datasheet and I've lost the link. I thought someone might be able to point me back at a similar device, but it looks like they're rare. Chris -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.17 (GNU/Linux) iEYEARECAAYFAk4EOcYACgkQXUF6hOTGP7cV9wCgiiCp6ChgbYsCsDeDBQxHK2TW bI0An2eXe1SkFpTourPkohWoVyQFTLer =ukA2 -----END PGP SIGNATURE-----

>> Do any of you know of ICs one can buy to perform cryptographic >> primitives (block ciphers, hashes, etc.)? I looked around Digikey in >> the categories that seemed reasonable, but couldn't seem to find any >> such devices. I know I've heard of one before (even seem its >> datasheet), but I can't seem to find the part now. > You may be thinking about the 8048 chip that Intel programmed to do DES encoding/decoding. I believe it was real slow (equivalent to around 100 baud IIRC) and required all sorts of clearances to be allowed to purchase it. I think I may have a pottle of them in my e-dungeon. Maybe 10 or 20 ? Acquired as part of a very large auction lot long ago. I think ITAR may not have even been a gleam in anyone's eye back then.

On Fri, 2011-06-24 at 00:16 -0700, Christopher Head wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > On Tue, 14 Jun 2011 23:13:09 -0700 > Christopher Head <headchKILLspamgmail.com> wrote: > > > Do any of you know of ICs one can buy to perform cryptographic > > primitives (block ciphers, hashes, etc.)? I looked around Digikey in > > the categories that seemed reasonable, but couldn't seem to find any > > such devices. I know I've heard of one before (even seem its > > datasheet), but I can't seem to find the part now. > > Thanks for the pointers everyone; it seems you haven't seen the kind of > thing I was looking for. Someone showed me a chip that would just > accept data on an input and produce a hash of the data on the output, > but I can't figure out who showed me the datasheet and I've lost the > link. I thought someone might be able to point me back at a similar > device, but it looks like they're rare. You sure it wasn't just a programmed FPGA/CPLD/PLD? Pretty trivial thing to do in one, I'd say that (and the fact that there are so many forms of encryption) is why a standalone chip is so rare. TTYL

On Fri, 2011-06-24 at 07:43 -0700, Harold Hallikainen wrote: > > > > You sure it wasn't just a programmed FPGA/CPLD/PLD? Pretty trivial thing > > to do in one, I'd say that (and the fact that there are so many forms of > > encryption) is why a standalone chip is so rare. > > > > I think the latest Power PC chips have encryption/decryption stuff in > them. Another guy here is using that in a project. Yes, quite a few chips have engines inside them as a subblock. But for discrete chips I see less purpose. It's just one of those functions that leads itself to either be part of the a larger function chip, or something added as part of a programmable design. TTYL

Some of the Power PC series have the ability to run encrypted code (decrypt and execute on the fly) but I don't think those modules allow you to encrypt/decrypt general data. Are you talking about some other additional function of the PPC family? Sean On Fri, Jun 24, 2011 at 10:43 AM, Harold Hallikainen <.....haroldKILLspam.....hallikainen.org> wrote: {Quote hidden}> >> >> You sure it wasn't just a programmed FPGA/CPLD/PLD? Pretty trivial thing >> to do in one, I'd say that (and the fact that there are so many forms of >> encryption) is why a standalone chip is so rare. >> > > I think the latest Power PC chips have encryption/decryption stuff in > them. Another guy here is using that in a project. > > Harold > >

> Some of the Power PC series have the ability to run encrypted code > (decrypt and execute on the fly) but I don't think those modules allow > you to encrypt/decrypt general data. Are you talking about some other > additional function of the PPC family? > > Sean I'm only on the periphery of that project, but I thought I heard that the PPC would handle encryption/decryption in hardware. It's being used in a digital cinema application. Again, I'm not directly involved in that project, so I may have misheard. Harold -- FCC Rules Updated Daily at http://www.hallikainen.com - Advertising opportunities available

maybe dedicated chips can still not compete with GPU given that they have 1600 parallel stream processors inside that apparently can be programed to create those hashes... I suppose unless you can integrate more than that number on an FPGA (because I doubt you could outclock it) the GPU will be Hash/time Hash/price winner! On Fri, Jun 24, 2011 at 2:35 PM, Harold Hallikainen <EraseMEharoldspam_OUTTakeThisOuThallikainen.org>wrote: {Quote hidden}> > > Some of the Power PC series have the ability to run encrypted code > > (decrypt and execute on the fly) but I don't think those modules allow > > you to encrypt/decrypt general data. Are you talking about some other > > additional function of the PPC family? > > > > Sean > > I'm only on the periphery of that project, but I thought I heard that the > PPC would handle encryption/decryption in hardware. It's being used in a > digital cinema application. Again, I'm not directly involved in that > project, so I may have misheard. > > Harold > > > -- > FCC Rules Updated Daily at http://www.hallikainen.com - Advertising > opportunities available! >