Q: SCA1000 is which OEM'd product?

Torsten Kirschner

unread,

Dec 21, 2003, 6:04:18 PM12/21/03

to

Hi,

does anyone happen to know what OEM product - if any - the
Sun Crypto Accelerator 1000 (SCA1000) is?

You know, like certain HBAs really are re-branded LSI HBAs,
and so on.

I am just trying to figure out whether or not this card
is supported by some security SDKs.

best regards
Torsten

Axel Neumann

unread,

Dec 22, 2003, 8:22:07 AM12/22/03

to

Hi,

AFAIK it is based on a Broadcom chip.

HTH,

Axel Neumann

Torsten Kirschner

unread,

Dec 22, 2003, 10:36:45 AM12/22/03

to

"Axel Neumann" <Axel.N...@epost.de> skrev i melding
news:bs6r6b$a6711$1...@ID-55633.news.uni-berlin.de...

Thanks. I am still hoping for an answer along the lines of
"it's actually a nCipher nFast or Rainbow CryptoSwift PCI". ;-)

You're problably right about the Broadcom chip, though. I'll
know for sure in january when I'll shutdown one of our machines
and take a closer look at the actual board.

The SCA1000's performance not just for "SSL" in general but lower
level stuff like the indivual algorithms (3DES and so forth) is
actually rather cool. We've measured a perfomance gain of over
two orders of magnitude running simple tests with free OpenSSL
and XMLSec (http://www.aleksey.com/xmlsec/) over commercial grade
Java implementations.
What I am looking for now is the best of both worlds, i.e.
using the SCA1000 from certain Java SDKs.

A kind soul at Sun has already pointed out the standard PKCS#11
interface, in case anything else fails.

Paul Eggert

unread,

Dec 22, 2003, 3:38:09 PM12/22/03

to

At Mon, 22 Dec 2003 16:36:45 +0100, "Torsten Kirschner" <torsten....@sandbox.no> writes:

> "Axel Neumann" <Axel.N...@epost.de> skrev i melding

>> AFAIK it is based on a Broadcom chip.
>
> Thanks. I am still hoping for an answer along the lines of
> "it's actually a nCipher nFast or Rainbow CryptoSwift PCI". ;-)

It's a Broadcom BCM5821, which is descended from the Bluesteel 5501.
(You'd know this stuff if you were running OpenBSD; they document it
better than Sun does. :-)

Scott Howard

unread,

Dec 27, 2003, 8:10:43 AM12/27/03

to

In comp.sys.sun.hardware Torsten Kirschner <torsten....@sandbox.no> wrote:
> does anyone happen to know what OEM product - if any - the
> Sun Crypto Accelerator 1000 (SCA1000) is?

It's not. The card itself is Sun designed, although as someone else has
pointed out it's based on a 3rd-party chip.

> You know, like certain HBAs really are re-branded LSI HBAs,
> and so on.

The original sun crypto card, the SCA1, was a re-badged Rainbow card, but
that's the only OEMed one.

Scott

Torsten Kirschner

unread,

Dec 27, 2003, 9:35:17 AM12/27/03

to

Scott Howard wrote:

> In comp.sys.sun.hardware Torsten Kirschner <torsten....@sandbox.no>
> wrote:
>> does anyone happen to know what OEM product - if any - the
>> Sun Crypto Accelerator 1000 (SCA1000) is?
>
> It's not. The card itself is Sun designed, although as someone else has
> pointed out it's based on a 3rd-party chip.

I see. Thanks for the confirmation. Well, I guess I'll just have to bite
the bullett and try to understand the OpenSSL and PKCS#11 APIs then.

As it happens, the C/C++ library most promising for what I am trying to
achieve, XMLSec from http://www.aleksey.com/xmlsec/, supports OpenSSL as
the crypto-engine. Thus I presume the same goes for Sun's SCA branding
of it.

As a last stab at it: how about the SCA4000, then? Is that of Sun's own
making as well?

best regards
Torsten

Scott Howard

unread,

Dec 28, 2003, 3:03:39 AM12/28/03

to

In comp.unix.solaris Torsten Kirschner <torsten....@sandbox.no> wrote:
> As a last stab at it: how about the SCA4000, then? Is that of Sun's own
> making as well?

Yup. Everything except the SCA1

Scott