Welcome to Apyfal’s Documentation!¶
Apyfal is a powerful and flexible toolkit that enables you to use FPGA  accelerated functions.”
Some reasons to use Apyfal :
- Apyfal provides an abstraction layer to use the power of FPGA accelerated function in a hybrid multi-cloud environment.
- The configuration and the provisioning is generated for the FPGA cloud context.
- Apyfal can perform acceleration directly on cloud storage objects.
- Don’t like Python ? Use the REST API and generate a client in any language.
All the accelerated functions
Apyfal provides a variety of accelerated functions.
Browse our Distribution platform, to discover them.
Basic Python code example
Accelerator API is easy to use and only needs a few lines of codes to instantiate an accelerator and its host and process data:
import apyfal # Choose and initialize an accelerator with apyfal.Accelerator(accelerator='my_accelerator') as myaccel: # Start and configure accelerator myaccel.start() # Process data using FPGA accelerated function myaccel.process(src='/path/myfile1.dat', dst='/path/result1.dat') myaccel.process(src='/path/myfile2.dat', dst='/path/result2.dat')
- Configuration of cloud host environment
- Remote or local execution
- Cloud storage direct access from accelerator
Limitations in remote mode
- Max input data size is limited to 30GB when using client local file.
- Timeout will occur if a request takes more than 900s.
- Getting Started
- Advanced use
- FAQ and Troubleshooting