Abstract
The LCLS2 Free Electron Laser (FEL) will generate x-ray pulses to beamline experiments at up to 1 Mhz. These experimentals will require new ultra-high rate (UHR) detectors that can operate at rates above 100 kHz and generate data throughputs upwards of 1 TB/s, a data velocity which requires prohibitively large investments in storage infrastructure. Machine Learning has demonstrated the potential to digest large datasets to extract relevant insights, however current implementations show latencies that are too high for real-time data reduction objectives. SLAC has endeavored on the creation of a software framework which translates MLs structures for deployment on Field Programmable Gate Arrays (FPGAs) deployed at the Edge of the data chain, close to the instrumentation. This framework leverages Xilinx’s HLS framework presenting an API modeled after the open source Keras interface to the TensorFlow library. This SLAC Neural Network Library (SNL) framework is designed with a streaming data approach, optimizing the data flow between layers, while minimizing the buffer data buffering requirements. The goal is to ensure the highest possible framerate while keeping the maximum latency constrained to the needs of the experiment. Our framework is designed to ensure the RTL implementation of the network layers supporting full re-deployment of weights and biases without requiring re-synthesis after training. The ability to reduce the precision of the implemented networks through quantization is necessary to optimize the use of both DSP and memory resources in the FPGA. We currently have a preliminary version of the toolset and are experimenting with both general purpose example networks and networks being designed for specific LCLS2 experiments.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Pragmas are a standard C/C++ feature used to communicate information directly to the compiler.
- 2.
Classes can be passed by reference. This technique is used to make some of the weights and biases directly available at synthesis time in the Reservoir layer.
- 3.
A HLS stream is the standard interface used to stream data between layers. It behaves like a FIFO. For technical reasons, the initial and final streams must be an AXI stream.
- 4.
The name of the class of the previous layer’s destination stream is well-defined, for example, presuming the previous layer parameter definition is layer2: :Parameters, then layer2: :Parameters: :DstStream.
References
Smith, T.F., Waterman, M.S.: Identification of common molecular subsequences. J. Mol. Biol. 147, 195–197 (1981). https://doi.org/10.1016/0022-2836(81)90087-5
May, Patrick, Ehrlich, Hans-Christian., Steinke, Thomas: ZIB structure prediction pipeline: composing a complex biological workflow through web services. In: Nagel, Wolfgang E.., Walter, Wolfgang V.., Lehner, Wolfgang (eds.) Euro-Par 2006. LNCS, vol. 4128, pp. 1148–1158. Springer, Heidelberg (2006). https://doi.org/10.1007/11823285_121
Foster, I., Kesselman, C.: The Grid: Blueprint for a New Computing Infrastructure. Morgan Kaufmann, San Francisco (1999)
Czajkowski, K., Fitzgerald, S., Foster, I., Kesselman, C.: Grid information services for distributed resource sharing. In: 10th IEEE International Symposium on High Performance Distributed Computing, pp. 181–184. IEEE (2001). https://doi.org/10.1109/HPDC.2001.945188
Foster, I., Kesselman, C., Nick, J., Tuecke, S.: The physiology of the grid: an open grid services architecture for distributed systems integration. Technical report, Global Grid Forum (2002)
National Center for Biotechnology Information. http://www.ncbi.nlm.nih.gov
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Herbst, R. et al. (2022). Implementation of a Framework for Deploying AI Inference Engines in FPGAs. In: Doug, K., Al, G., Pophale, S., Liu, H., Parete-Koon, S. (eds) Accelerating Science and Engineering Discoveries Through Integrated Research Infrastructure for Experiment, Big Data, Modeling and Simulation. SMC 2022. Communications in Computer and Information Science, vol 1690. Springer, Cham. https://doi.org/10.1007/978-3-031-23606-8_8
Download citation
DOI: https://doi.org/10.1007/978-3-031-23606-8_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-23605-1
Online ISBN: 978-3-031-23606-8
eBook Packages: Computer ScienceComputer Science (R0)