FAQ

1. General Questions

What is Corazon AI?

Corazon-AI is an edge AI platform built on Zynq Ultrascale+ MPSoC device integrated with
Xilinx Vitis-AI stack enabling instantaneous real-time decisions on the edge.

Is Corazon-AI a Development platform or a Solution platform?

Corazon-AI is a solution platform where one can directly deploy use cases with minimum hardware design effort.

Corazon AI and ZCU104, are they the same?

No.
Corazon-AI is a solution platform and it is designed such that the developer can develop
their end application directly on this platform without the need for redesigning the
hardware.
ZCU104 platform is a development platform. After evaluation, the user can redesign the
custom hardware for use cases.

Is there any demo available?

Yes. We have demos available for
face recognition,
people counting,
social distancing,
Object detection, etc.

How can I run my applications on Corazon-AI?

Please refer to our software user guide to run the applications.

Are there any pre-built binaries available for Corazon-AI?

The Corazon AI deliverables have pre-built applications and their source code.
Refer to the software user guide to run the pre-built applications.
Also, the pre-built application source code can be modified as per customization
requirements.

2. Hardware

How do I access the hardware user guide for Corazon-AI?

The deliverable package will have the hardware user guide which will be shipped along with our Corazon AI platform.
Also, the user can request the same by writing an email to the mail ID
Fpga-design@iwavesystems.com
How can I get a loaner board?

Users can request the same by writing the email to the below mail ID.
Fpga-design@iwavesystems.com

How do I connect my hardware peripherals to Corazon AI?

Please refer to our hardware user guide for the peripheral connection procedure.

How do I power ON the Corazon-AI platform?

Please refer to our Quickstart guide for the Power ON procedure.

How to connect the debug port of the Corazon AI platform to the host system?

Please refer to our Quickstart guide for the debug port connection procedure.

Is debug console available in Corazon AI?

Yes. The debug console cable is part of deliverables for the Corazon AI Engineering version
platforms

Is WiFi and Bluetooth supported

Yes. Please refer to our product hardware configuration for the exact Hardware platform.

Is LTE supported

Yes. Please refer to our product hardware configuration for the exact Hardware platform.

3. FPGA

Can I modify the FPGA design?

Yes, it is possible to add additional interfaces such as sensors/other interfaces.
Please refer to our FPGA user guide to modify the FPGA design.

Can I develop an only application without modifying the Peta-Linux and FPGA design

We provide pre-built binaries enabled with library and Packages required for a developer to
develop the application.
Petalinux or FPGA design customization is required when the pre-build binaries are not
meeting the required application specification.

4. Software

How do get the software deliverables?

The deliverable package will be shipped along with our Corazon AI platform.
Also, the user can request the same by writing the email to the below mail ID.
fpga-design@iwavesystems.com

How is the host system configuration used for the development?

Please refer to our software user guide for the Host system configuration

How to run my applications/modify the pre-built applications

The Corazon AI deliverables have pre-built applications and their source code.
Refer to the software user guide to run the pre-built applications.
Also, the pre-built application source code can be modified as per user customization.

Are there any pre-built binaries?

Yes. The Corazon AI deliverable package has pre-built Neural network Models and sample
application.
Please refer to our software user guide to run the sample applications.

Can I modify Peta-Linux?

Yes. It is possible to add additional interfaces such as sensors/other interfaces.
Please refer to our software user guide to modify the Peta-Linux design.

Can I run custom applications?

Yes, please refer to our Release note to confirm the required libraries or package added already. In case of the additional package for a custom application, please refer to our software
user guide to add packages in Peta-Linux.

Can I run custom models?

Yes. The user can deploy their custom models, CNN-based models, on the platform.
Deep Neural Network Features and parameters support the DPU.

Can I develop an only application without modifying the Peta-Linux and FPGA design

We provide pre-built binaries enabled with library and Packages required for a developer to develop the application.
Petalinux or FPGA design customization is required when the pre-build binaries are not
meeting the required application specification

Is there any cloud connections supported?

Yes, so far the platform supported the three main stream cloud platforms AWS, IBM, and Azure.
For the configuration support, please write an email to the below email id.
fpga-design@iwavesystems.com

What is the boot media supported by default?

By default, the platform supports emmc booting media.

Can I run my applications from a USB memory stick?

Yes. For the flexibility development testing, the platform supports multiple USB3.0 slots to connect the pen drive devices. Users can use these slots to connect the pen drive and run their applications.

Is NFS supported in Corazon AI?

Yes. Refer to our Software user guide to validate NFS.

Can I run my applications via NFS

Yes, user can run the application via NFS

Can I run my application without connecting display?

Yes. The X11 forwarding display and Gstreamer options are available for the user to stream the output frames on the host PC via ethernet.

Can I steam my processed output to the host system via ethernet?

Yes. The X11 forwarding display and Gstreamer options are available for the user to stream the output frames on the host PC via ethernet.

5. Cloud

Can I send processed data to the cloud?

Yes, the user can use the MQTT-based application to send the data to the cloud.

Can I send my processed data to the cloud via WiFi or Ethernet or LTE?

It has been tested using Wifi & Ethernet connectivity.