http://www.mc.com/products/view/index.c ... ype=boards
More than 179 GFLOPS in PCI Express accelerator card
VERSUS
2.6 gigaflop peak performance on a scope professional card.
Oh, the possibilities.
/* drool */
$8k for the board plus an amount x for the software, let alone you'll have to implement all and any audio processing yourself...Pricing for the Mercury Cell Accelerator Board starts at US$7,999 for a single board, with discounts available on higher volumes.
...Mercury's systems are also used in state-of-the-art medical diagnostic imaging devices including MRI, PET, and digital X-ray, and in semiconductor imaging applications including photomask generation and wafer inspection. ...
just to find out that your 'data' dripples in at (say) 96khz, rendering most processing capabilities of such a device to a waste of resources.
Even ultraound imagers (which afaik have at least 10 times the data rate of digital audio) aren't mentioned in the list above. The machines quoted are $ million investments
Mercury certainly is an interesting company with fascinating products, but regarding to audio it's rather off topic.
I don't see any 'possibilities at all.
Who feels challenged can easily start with Analog Devices TigerSharcs (regarding sheer processing power), at least their DSP software has a huge audio library.
In any case you'd need the VisualDSP sdk for $3.5k - maybe 2 Scope professionals are a more effective investment...
Afaik there's a $2.5k guitar fx processor available, based on that type of chip (1), but it doesn't sound much different (if at all) to Dynatube to my ears (according to demo sounds).
cheers, Tom
<font size=-1>[ This Message was edited by: astroman on 2006-10-04 04:48 ]</font>
another idea would be to not bring out completly new hardware, but to make the existing one more affordable....
For example a Scope booster Professional card for 500€ without software would be a nice buy and I would instantly go for one.
just without any bundled software. I think the hardware should be quite cheap now.
CW should make their main money with the plug-ins (this is where the main know-how is behind)and make the dsp power more affordable.
my 2 cent
<font size=-1>[ This Message was edited by: hifiboom on 2006-10-04 09:41 ]</font>
For example a Scope booster Professional card for 500€ without software would be a nice buy and I would instantly go for one.
just without any bundled software. I think the hardware should be quite cheap now.
CW should make their main money with the plug-ins (this is where the main know-how is behind)and make the dsp power more affordable.
my 2 cent
<font size=-1>[ This Message was edited by: hifiboom on 2006-10-04 09:41 ]</font>
maybe hacking a PSP is the way to go then... 
I've read it will be heavily sponsored to meet street prices for xmas sales and such.
The Cell Cpu alone is said to cost Sony(!) $230
the Cell's architecture isn't that different from a Scope board, but of course in a 'modern' design with a controlling CPU on-chip together with 8 vector units and a ton of on-chip memory for each of them.
obviously Sony itself is trying to place it as a 'computer' instead of 'just' a game console, so if the box isn't entirely closed... and some cross-compiler exists... and someone has the time I don't have...
looks promising indeed, since it doesn't need all the super-performance IO of the Mercury board (no positron-emission-tomography, no radio frequency digitization), a PSP is the way to go.
If it doesn't work out as expected in the synth and sound domain, it will at least be an error with a high entertainment factor
cheers, Tom
I've read it will be heavily sponsored to meet street prices for xmas sales and such.
The Cell Cpu alone is said to cost Sony(!) $230
the Cell's architecture isn't that different from a Scope board, but of course in a 'modern' design with a controlling CPU on-chip together with 8 vector units and a ton of on-chip memory for each of them.
obviously Sony itself is trying to place it as a 'computer' instead of 'just' a game console, so if the box isn't entirely closed... and some cross-compiler exists... and someone has the time I don't have...
looks promising indeed, since it doesn't need all the super-performance IO of the Mercury board (no positron-emission-tomography, no radio frequency digitization), a PSP is the way to go.
If it doesn't work out as expected in the synth and sound domain, it will at least be an error with a high entertainment factor
cheers, Tom