Why doesn't Sun release the crypto module of the OpenSPARC? Competition?

[David G. Koontz]

Lawerence Spracklen's Blog:
http://blogs.sun.com/sprack/entry/detailed_t2_crypto_info

  Detailed T2 crypto info

  Very detailed info on the UltraSPARC T2 cryptographic accelerators can be
  found here on the OpenSPARC website (the pertinent info can be found in
  chapter-21 of the doc)

  Posted on: Sep 11, 2007

With a rebuttal that the Ch 21 in the document found there contained the PCI
Express Interface Unit:

  Unfortunately, it looks like the accelerator details have been removed :-(
  The SPU is not technically part of OpenSPARC....

  Posted by Lawrence Spracklen on November 05, 2007 at 11:39 AM PST #

There's the aspect of competition.  The on core crypto gives one heck of a
competitive edge for networking applications, and performance figures found
on Dr. Spracklen's site show that the crypto stream processors across the
CMT can keep up with the 10G Ethernet ports.  I can't see them giving a
potential competitor everything needed to compete directly.  It'd be
reminiscent of IBM and Amdahl clones, captive markets and margins for
hardware threatened as easily as National's memory boards.  I'm sure Sun is
wiley enough to have some key patents, too.  A case of encouraging help to
enlarge the ecosystem, but not empowering direct competition.  They don't
mind if you develop more markets, after all Sun can play there, too.

I've also wondered if a reason they didn't release it is because they bought
the 'IP' from someone.  There are other instances - parts of the System on a
Chip.  In the OpenSPARC T2 System on a Chip Micro Architecture pdf there is
a disclaimer on page 3:

  Note ? OpenSPARC T2 currently does not include PCI-Express and 10Gigabit
  Ethernet design implementation due to current legal restrictions.
  Equivalent models may be available in the subsequent releases of OpenSPARC
  T2.

If the real reason is competition, it's always nice to have a good excuse,
too.

---------------------------------------------------------------------
The Cryptography Mailing List
Unsubscribe by sending "unsubscribe cryptography" to majordomo at metzdowd.com