Hello dears.
im beginer in opencl, and i want to run complex QR decomposition opencl kernel in altera(intel) cyclone-v FPGA and GPU for part of my master thesis.
my FPGA does not have enough hardware resource to run viennacl QR kernel compeletly. for this reason i want to run some part of oepncl kernel in FPGA.
i could not find complex opencl kernel of QRD with .cl format.
is viennacl kernel open source compeletly?
how can i find compelex QRD opencl kernel?
help me please.
Thanks.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
QR decomposition uses not a single kernel, but a sequence of different kernels. Thus, you won't find a single kernel or .cl-file that implements QR decomposition (btw: yes, ViennaCL is fully open source). As such, you would have to make sure that all kernels are executed on the device. If only some kernels are run on the FPGA and others not, you will likely get poor performance because of the data transfers from/to the FPGA.
Best regards,
Karli
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello dears.
im beginer in opencl, and i want to run complex QR decomposition opencl kernel in altera(intel) cyclone-v FPGA and GPU for part of my master thesis.
my FPGA does not have enough hardware resource to run viennacl QR kernel compeletly. for this reason i want to run some part of oepncl kernel in FPGA.
i could not find complex opencl kernel of QRD with .cl format.
is viennacl kernel open source compeletly?
how can i find compelex QRD opencl kernel?
help me please.
Thanks.
Hi Mahmood,
sorry for the late reply :-/
QR decomposition uses not a single kernel, but a sequence of different kernels. Thus, you won't find a single kernel or .cl-file that implements QR decomposition (btw: yes, ViennaCL is fully open source). As such, you would have to make sure that all kernels are executed on the device. If only some kernels are run on the FPGA and others not, you will likely get poor performance because of the data transfers from/to the FPGA.
Best regards,
Karli