1994-08-30 - Re: DSPs

Header Data

From: jdd@aiki.demon.co.uk (Jim Dixon)
To: eb@comsec.com
Message Hash: 4d66d720bd47d75c5d4bbcb109aa9006497ab314d528b53ce49f07ade550d304
Message ID: <88@aiki.demon.co.uk>
Reply To: N/A
UTC Datetime: 1994-08-30 15:16:06 UTC
Raw Date: Tue, 30 Aug 94 08:16:06 PDT

Raw message

From: jdd@aiki.demon.co.uk (Jim Dixon)
Date: Tue, 30 Aug 94 08:16:06 PDT
To: eb@comsec.com
Subject: Re: DSPs
Message-ID: <88@aiki.demon.co.uk>
MIME-Version: 1.0
Content-Type: text/plain


In message <199408292302.QAA02577@comsec.com> Eric Blossom writes:
> > The Motorola DSP96002 does an integer multiply in 2 or 3 clocks, so a
> > 33 MHz device does 11 million multiplies (and moves) a second. The
> > chip costs about $50.
> 
> The 96002 is a floating point part.  Last time I checked it cost
> several hundred dollars.  I suspect that you were refering to the
> 56001/2 family (which does cost something like $50).

I checked the 56001 data book, but it does 24 bit integer multiplies,
not 32, so I didn't quote them.  The figures I gave are for a 32-bit
integer multiply, right out of the 96002 data book.

I haven't checked 96002 prices for a couple of years.  When I last
checked it was something like 200 pounds ($300) in small quantities.
It has now been outclassed by the C40 and several other very good
DSP chips and time has passed, so I think that in production
quantities the price would at least be under the $100 mark.
-- 
Jim Dixon





Thread