OCLADock-FPGA: OpenCL Accelerated Molecular Docking on FPGAs
=============================================================

<img src="logo.png" width="200">

# Description

OpenCL task-parallel version of AutoDock4 for FPGAs. 

It achieves speed-ups of up to 2.7x over the original single-threaded AutoDock 4.2 on CPU.

# Development branches

The master branch points to the latest OCLADock-FPGA version: [lga-aoc160](https://git.esa.informatik.tu-darmstadt.de/docking/ocladock-fpga/tree/lga-aoc160).

The other branches are under regular development & optimization. Their labels follow this convention: `<ALGORITHM>-<TOOL>`:

* [ga-aoc160](https://git.esa.informatik.tu-darmstadt.de/docking/ocladock-fpga/tree/ga-aoc160): Genetic Algorithm (traditional) - Altera OpenCL Compiler v16.0
* [lga-intel171](https://git.esa.informatik.tu-darmstadt.de/docking/ocladock-fpga/tree/lga-intel171): Lamarckian Genetic Algorithm - Intel FPGA SDK for OpenCL v17.1
* [lga-sdx174](https://git.esa.informatik.tu-darmstadt.de/docking/ocladock-fpga/tree/lga-sdx174): Lamarckian Genetic Algorithm - Xilinx SDAccel v2017.4

# Development tools

Specific information of the utilized OpenCL tools can be found in the respective README.md for each development branch.

Initialization scripts automate the tool setup-process. E.g.:

```zsh
% source init_aoc160_gidel.sh
```

## Important commands

See rules in [Makefile](./ofdock_taskpar_alt/Makefile).

**OpenCL kernels compilation**

```zsh
make keremu
```

**Design emulation**

```zsh
make emu
```

**Estimation & optimization reports generation**

```zsh
make kerrpt
```

**FPGA building**

```zsh
make hw
```

**FPGA building for hardware profiling**

```zsh
make hw-prof
```

**FPGA evaluation**

```zsh
make exe
```

# Documentation

For more details, go to the [Documentation](doc/readme/home.md).

# Bibliographic information for citing OCLADock-FPGA

Leonardo Solis-Vasquez and Andreas Koch. 2018. A Case Study in Using OpenCL on FPGAs: Creating an Open-Source Accelerator of the AutoDock Molecular Docking Software. In Proceedings of the 5th International Workshop on FPGAs for Software Programmers (FSP), 2018. 

# License

This project is licensed under the GNU GPLv2 - see the [LICENSE](LICENSE) file for details.

# Related projects

A data-parallel OpenCL implementation of AutoDock4 for CPUs and GPUs can be found in the [OCLADock](https://git.esa.informatik.tu-darmstadt.de/docking/ocladock) repository .