@@ -23,24 +23,21 @@ This document provides instructions on how to execute the provided FPGA-accelera
The application developed as a demonstrator for EDRA is based on the [RAxML application](https://github.com/stamatak/standard-RAxML). The application has been modified according to the Decoupled Access Execute paradigm developed within the EDRA project and an FPGA accelerator has been developed to perform the most demanding computational task. The FPGA-accelerated application is designed to be deployed on an Amazon F1 instance and for demonstration purposes it includes a small data set as well as a software component that is meant to verify that the results obtained be the original application and the accelerated one are equivalent.
The accelerated application is provided in a binary form. The software components are the compiled source code running on the host processor and the FPGA accelerator is in the form of an AWS-compatible binary file containing the FPGA bitstream and all other components required to execute the application. A small dataset is also provided for a sample run. Source code for the software application and the accelerator can be provided upon request.
The accelerated application is provided in binary form. The software components are the compiled source code running on the host processor and the FPGA accelerator is in the form of an AWS-compatible binary file (.awsxclbin) containing the FPGA bitstream and all other components required to execute the application. A small dataset is also provided for a sample run. Source code for the software application and the accelerator can be provided upon request.
To execute the accelerated application
To execute the accelerated application, access to an AWS F1 instance is required. The following steps describe the process to prepare a proper environment on AWS that can be used to deploy the application and the instructions to actually execute it on the AWS instance. In case you have already setup AWS accounts and deployed F1 instances, feel free to skip the initial steps and proceed directly to the execution of the application.
* SDAccel is a complete development environment for applications accelerated using Xilinx FPGAs
* It leverages the OpenCL heterogeneous computing framework to offload compute intensive workloads to the FPGA
* The accelerated application is written in C/C++, OpenCL or RTL with OpenCL APIs
* Once you complete this quick starting example, see the [SDAccel GUI Guide](./docs/README_GUI.md) to access the fully integrated Eclipse-based environment with built-in debug, profiling and performance analysis tools.
It should be mentioned that since the current repository provides the application in binary form and not source code, you are not required to setup a development environment nor obtain third party licenses. The demonstration application is provided for free from the the EDRA project, however have in mind that there are associated usage fees related to the use of EC2/F1 instances from Amazon.
<aname="prerequisites"></a>
# Prerequisites
<aname="iss"></a>
## AWS Account, F1/EC2 Instances, On-Premises, AWS IAM Permissions, AWS CLI and S3 Setup (One-time Setup)
## AWS Account, F1/EC2 Instances, AWS IAM Permissions, AWS CLI and S3 Setup (One-time Setup)
If you do not already have an AWS account, you should follow these steps to establish an account and setup a proper F1 instance for the demo application.
*[Setup an AWS Account](https://aws.amazon.com/free/)
* Launch an instance using the [FPGA Developer AMI](https://aws.amazon.com/marketplace/pp/B06VVYBLZZ) which comes pre-installed with SDAccel and required licenses.
*You may use this F1 instance to [build your host application and Xilinx FPGA binary](#createapp), however, it is more cost efficient to either:
* Launch an instance using the [FPGA Developer AMI](https://aws.amazon.com/marketplace/pp/B06VVYBLZZ). which comes pre-installed with SDAccel and required licenses.
*The demonstration application has been developed using Xilinx tools version 2018.3, therefore be careful to select a compatible AMI version. It should be marked as <i>v1.6.0-v1.6.X (Xilinx Vivado/SDx 2018.3)</i>
* Launch the [FPGA Developer AMI](https://aws.amazon.com/marketplace/pp/B06VVYBLZZ) on a compute EC2 instance, with a minimum of 30GiB RAM), **OR**
* Follow the [On-Premises Instructions](../hdk/docs/on_premise_licensing_help.md) to purchase and install a license from Xilinx.
* Setup AWS IAM permissions for creating FPGA Images (CreateFpgaImage and DescribeFpgaImages). [EC2 API Permissions are described in more detail](http://docs.aws.amazon.com/AWSEC2/latest/APIReference/ec2-api-permissions.html). It is highly recommended that you validate your AWS IAM permissions prior to proceeding with this quick start. By calling the [DescribeFpgaImages API](../hdk/docs/describe_fpga_images.md) you can check that your IAM permissions are correct.
