Intel® Distribution of OpenVINO™ Toolkit
Community assistance about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all aspects of computer vision-related on Intel® platforms.

Multiple FPGAs with OpenVINO

jliu32
Beginner
482 Views

Suppose I have more than one Arria 10 Development Kits installed on the motherboard, how can I select which board to use for a given inference request?

0 Kudos
4 Replies
Monique_J_Intel
Employee
482 Views

Hi James,

To use multiple FPGAs you can do so by specifying the KEY_DEVICE_ID and the number. In this example I have 2 FPGAs:

ExecutableNetwork  excutable_network1 = plugin.LoadNetwork(network, {InferenceEngine::PluginConfigParams::KEY_DEVICE_ID, "0"});
ExecutableNetwork  excutable_network1 = plugin.LoadNetwork(network, {InferenceEngine::PluginConfigParams::KEY_DEVICE_ID, "1"});

Then from here you can create the infer request for each FPGA respectively.

Kind Regards,

Monique Jones

0 Kudos
jliu32
Beginner
482 Views

Hi Monique,

   Thanks for the prompt reply. Does device id "x" in OpenVINO map to "aclx" seen by aocl then? 

Regards,

James

0 Kudos
Monique_J_Intel
Employee
482 Views

Hi James,

Yes, that is correct.

Kind Regards,

Monique Jones

0 Kudos
Shawn_S_Intel
Employee
482 Views

The code that Monique specified appears to require double curly braces for compilation:

ExecutableNetwork executable_network = plugin.LoadNetwork(network, {{InferenceEngine::PluginConfigParams::KEY_DEVICE_ID, "0"}});

 

0 Kudos
Reply