User
Applications
|
Seismic
Imaging Acceleration
Our FPGA PCI
acceleration cards enable
you to dramatically improve the performance of your seismic imaging
applications while simultaneously reducing the cost, size, and power
requirements of your computing clusters. Our cards enable you to
harness the power of FPGAs by creating heterogeneous multi-core
systems in an FPGA with each core customized to perform its task
optimally. Clusters of cores are connected in a data flow arrangement
to support high data bandwidth and the clusters are connected by
shared memory for flexibility.
 You will benefit
from
using our cards
through a lower total cost of ownership. Our FPGA PCI acceleration
cards are faster than servers, yet still have a power
savings advantage over them. Being more powerful also means that you
need less room for your cluster. You can now build powerful portable
systems that allow you to preview your data in the field or even at
sea.
Save on size, weight, and power by swapping servers for our FPGA
acceleration cards.
Put
the power of
our FPGA cards to work
for you now. Our Impulse
CTM to FPGA support
allows you to port your
seismic imaging applications to our cards quickly and productively.
Partition your code into communicating sequential processes,
which are similar to threads. Each of these processes can be turned
into a custom core on the FPGA or run as software on an embedded
processor. This partitioning allows the tools to extract the data
flow and high level opportunities for parallelism. The Impulse
CTM tools
then infer the low level parallelism or you can guide it with
compiler directives.
Our
multi-gigabit serial I/Os allow
efficient high-speed communication between our cards and disk. You
can directly connect cards inside the same host with our SATA/SAS
connectors or network them together by using the SFP connections. Low
latency DDR2 memory connected directly to the FPGA delivers the
Gigabytes you need to keep your application fed.
|
