Creamware has announced a multiple SHARC (yes, the Analog Device chip) board
that does signal processing and synthesis. Price and interfaces not yet
known.

Anyone know more about this?

In my opinion, the future of music synthesis is strictly software and no
hardware at all, but for several years it will not be possible to do thick
arrangements in low-latency real time using only software on personal
computers, so hardware acceleration makes good sense for the time being, if
someone does it right.

-----Original Message-----
From: Nicola Bernardini 
To: Csound mailing list 
Date: Thursday, April 09, 1998 1:11 PM
Subject: Re: ADI Csound SDK/Csounders swear !?!


>On Tue, 7 Apr 1998, khalid wrote:
>
>[snip]
>> Nicola Bernardini wrote:
>> >
>> >
>> > Why? If the specs where public, it would probably be so quick to
>> > build the *right* development tools for that card; oh well, who cares,
>> > there are so many other cards...
>> >
>> > Nicola
>>
>> I haven't noticed any sign that ADI is actually aiming at developing a
>> PCI card for the market. I guess they make these cards for developement
>> of apps that would then run in embedded systems. If you consider how
>> many people (buyers) actually care bout things like "LPC" or grain-
>> synthesis you might understand them. We software synthesis lovers are
>> still a minority that won't grow so quickly, I believe.
>
>I do perfectly understand that ADI is not interested. As a matter of
>fact, in another very clear and illuminating mail by Scotty Vercoe
>(to whom I must express my gratitude to have taken the time to explain
>so plainly and kindly what is going on), Scotty says:
>
>On Tue, 7 Apr 1998, Vercoe, Scotty wrote:
>
>[snip]
>> exciting development environment.  Keep in mind that Analog Devices is a
>> chip company, so doing software alone is a stretch!  Our Software and
>> Systems Technology Division led by Mike Haidar, is dedicated to
>> enhancing the value of ADI's chips with software for dedicated
>> applications.  Writing drivers for multiple platforms does not make
>^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> sense because we are not trying to make money off the development kit.
>^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> The goal of any kit from ADI is to provide customers and developers with
>> a system they can use to prototype products that use our chips.
>
>I find this perfectly sensible and reasonable. And I would think that
>ADI has all to be gained to keep the information concerning their
>software activities and writing public, so that all of us could contribute
in
>providing more platform implementations etc. etc.
>
>> Meanwhile, you could have a look at the MAAS/Highend-sound-driver pro-
>> ject. This board has a Motorola 56301 DSP on it. It is said to reach
>> 80.000 MFlops (the SHARC DSP is about 120.000). There are development
>> tools for the 56k family running under Linux (though I haven't used
>
>I am aware of the MAAS project and especially of the Highend-sound-driver
>and I am in contact with Guenter Geiger; I find them very courageous and
>I wonder if and when and for how much a card like the MAAS could be had...
>The point is: we need high diffusion cards to be able to buy them at
>a reasonable cost, and at the same time we need the info to be able
>to add linux support. Since the commercial guys do not release the info
>to avoid cloning of the cards, and linux/csound/whatever is a small, very
>small niche, we are stuck. This is why generic development boards like
>ADI's could come in handy. I must say that the last benchmark war on
>hardwares and OSes on this mailing list was pretty boring (I hope that
>is over now): everybody should use whatever works best for her/him and
>her/his pocket - but in order to do that, we must obtain support to be
>able to build drivers to their full extent etc.
>
>> them). B.t.w. - great work with the parser!!
>
>Thank you. I *really* appreciate any comments on that...
>
>nicb
>------------------------------------------------------------------------
>Nicola Bernardini
>E-mail: nicb@axnet.it
>
>Re graphics: A picture is worth 10K words -- but only those to describe
>the picture.  Hardly any sets of 10K words can be adequately described
>with pictures.
>
>