User
Applications
|
DSP
Acceleration
 By utilizing
the
embedded XtremeDSPTM
blocks, our PCI DSP Acceleration solutions deliver substantially more
signal processing power than competing products using DSP chips. By
pairing the XtreamDSPTM blocks with the embedded block RAMS,
our
PCI DSP Acceleration cards solve demanding problems such as
asynchronous resampling filters, polyphase filters, adaptive filters
and FFTs. The incredible amount of bandwidth available between the
XtreamDSPTM blocks and the block RAMs enables the FPGA to
solve DSP
problems that would be bandwidth limited on DSP chips.
Our
flexible
design environment allows
you create pipelines of DSP blocks that are closely matched to your
requirements. By wrapping DMA engines around pipelines of DSP
blocks, you create custom DSP processors that can be dedicated to one
task, or shared between many. You can have numerous small DSP
processors
for maximum flexibility, or you can build deep DSP pipelines for
maximum efficiency in solving your most demanding task. Think of it
as building a custom heterogeneous multi-core DSP system.
Our
Impulse
CTM support and reference
designs allow you to quickly implement a new algorithm and have it
solving your problems today. You can design and test your algorithm
on a PC, then move it to the FPGA when you are ready to use it in
your full solution.
The
pair of
embedded PowerPCs
inside
the FPGA on our cards allows you to run some of your code on them,
freeing up more of the FPGA resources for the high performance parts
of your application. By creating DSP processor pipelines, the code
on the PowerPCs can command your DSP processors to handle large
amounts of data while it deals with other chores. This combines the
power of customized FPGA hardware with the flexibility of
software.
|
