In many experiment-driven scientific domains, such as high-energy physics, material science, and cosmology, high data rate experiments impose hard constraints on data acquisition systems: collected data must either be indiscriminately stored for post-processing and analysis, thereby necessitating large storage capacity, or accurately filtered in real-time, thereby necessitating low-latency processing. Deep neural networks, effective in other filtering tasks, have not been widely employed in such data acquisition systems, due to design and deployment difficulties. We present an open source, lightweight, compiler framework, without any proprietary dependencies, OpenHLS, based on high-level synthesis techniques, for translating high-level representations of deep neural networks to low-level representations, suitable for deployment to near-sensor devices such as field-programmable gate arrays. We evaluate OpenHLS on various workloads and present a case-study implementation of a deep neural network for Bragg peak detection in the context of high-energy diffraction microscopy. We show OpenHLS is able to produce an implementation of the network with a throughput 4.8 $μ$s/sample, which is approximately a 4$\times$ improvement over the existing implementation
In this paper we present our contribution in terms of synchronization processor for a SoC design methodology based on the theory of the latency insensitive systems (LIS) of Carloni et al. Our contribution consists in IP encapsulation into a new wrapper model which speed and area are optimized and synthetizability guarantied. The main benefit of our approach is to preserve the local IP performances when encapsulating them and reduce SoC silicon area.
Packet parsing is a key step in SDN-aware devices. Packet parsers in SDN networks need to be both reconfigurable and fast, to support the evolving network protocols and the increasing multi-gigabit data rates. The combination of packet processing languages with FPGAs seems to be the perfect match for these requirements. In this work, we develop an open-source FPGA-based configurable architecture for arbitrary packet parsing to be used in SDN networks. We generate low latency and high-speed streaming packet parsers directly from a packet processing program. Our architecture is pipelined and ent…
Recent efforts for improving the performance of neural network (NN) accelerators that meet today's application requirements have given rise to a new trend of logic-based NN inference relying on fixed function combinational logic. Mapping such large Boolean functions with many input variables and product terms to digital signal processors (DSPs) on Field-programmable gate arrays (FPGAs) needs a novel framework considering the structure and the reconfigurability of DSP blocks during this process. The proposed methodology in this paper maps the fixed function combinational logic blocks to a set o…
Balancing (equalization) of latency in parallel paths in the pipelined data processing system is an important problem. Without that the data from different paths arrive at the processing blocks in different clock cycles, and incorrect results are produced. Manual correction of latencies is a tedious and error-prone work. This paper presents an automatic method of latency equalization in systems described in VHDL. The method is based on simulation and is portable between different simulation and synthesis tools. The method does not increase the complexity of the synthesized design comparing to …
In this work, we present a new approach to high level synthesis (HLS), where high level functions are first mapped to an architectural template, before hardware synthesis is performed. As FPGA platforms are especially suitable for implementing streaming processing pipelines, we perform transformations on conventional high level programs where they are turned into multi-stage dataflow engines [1]. This target template naturally overlaps slow memory data accesses with computations and therefore has much better tolerance towards memory subsystem latency. Using a state-of-the-art HLS tool for the …
For polar codes with short-to-medium code length, list successive cancellation decoding is used to achieve a good error-correcting performance. However, list pruning in the current list decoding is based on the sorting strategy and its timing complexity is high. This results in a long decoding latency for large list size. In this work, aiming at a low-latency list decoding implementation, a double thresholding algorithm is proposed for a fast list pruning. As a result, with a negligible performance degradation, the list pruning delay is greatly reduced. Based on the double thresholding, a low-…
Modern internet of things (IoT) devices leverage machine learning inference using sensed data on-device rather than offloading them to the cloud. Commonly known as inference at-the-edge, this gives many benefits to the users, including personalization and security. However, such applications demand high energy efficiency and robustness. In this paper we propose a method for reduced area and power overhead of self-timed early-propagative asynchronous inference circuits, designed using the principles of learning automata. Due to natural resilience to timing as well as logic underpinning, the cir…
UC Santa Barbara（加州大学圣塔芭芭拉分校）；Georgia Institute of Technology（佐治亚理工学院）
Despite the great success of High-Level Synthesis (HLS) tools, we observe several unresolved challenges: 1) the high-level abstraction of programming styles in HLS sometimes conceals optimization opportunities; 2) existing HLS tools do not provide flexible trade-off (Pareto) solutions among different objectives and constraints; 3) the actual quality of the resulting RTL designs is hard to predict. To address these challenges, we propose an end-to-end framework, namelyIronMan. The primary goal is to enable a flexible and automated design space exploration (DSE), to provide either optimal soluti…
LL-GNN: Low Latency Graph Neural Networks on FPGAs for High Energy Physics
Imperial College London（伦敦帝国理工学院）；Southeast University（东南大学）；AMD Adaptive and Embedded Computing Group (AECG) Labs（AMD自适应与嵌入式计算集团实验室）；European Organization for Nuclear Research (CERN)（欧洲核子研究中心）
Early scheduling algorithms usually adjusted the clock cycle duration to the execution time of the slowest operation. This resulted in large slack times wasted in those cycles executing faster operations. To reduce the wasted times multi-cycle and chaining techniques have been employed. While these techniques have produced successful designs, its effectiveness is often limited due to the area increment that may derive from chaining, and the extra latencies that may derive from multicycling. In this paper we present an optimization method that solves the time-constrained scheduling problem by t…
High-level synthesis (HLS) enables designers to customize hardware designs efficiently. However, it is still challenging to foresee the correlation between power consumption and HLS-based applications at an early design stage. To overcome this problem, we introduce HL-Pow, a power modeling framework for FPGA HLS based on state-of-the-art machine learning techniques. HL-Pow incorporates an automated feature construction flow to efficiently identify and extract features that exert a major influence on power consumption, simply based upon HLS results, and a modeling flow that can build an accurat…
California Institute of Technology（加州理工学院）；European Organization for Nuclear Research (CERN)（欧洲核子研究组织（CERN））；Imperial College London（帝国理工学院）
This paper presents novel reconfigurable architectures for reducing the latency of recurrent neural networks (RNNs) that are used for detecting gravitational waves. Gravitational interferometers such as the LIGO detectors capture cosmic events such as black hole mergers which happen at unknown times and of varying durations, producing time-series data. We have developed a new architecture capable of accelerating RNN inference for analyzing time-series data from LIGO detectors. This architecture is based on optimizing the initiation intervals (II) in a multi-layer LSTM (Long Short-Term Memory) …
Custom hardware accelerators for Deep Neural Networks are increasingly popular: in fact, the flexibility and performance offered by FPGAs are well-suited to the computational effort and low latency constraints required by many image recognition and natural language processing tasks. The gap between high-level Machine Learning frameworks (e.g., Tensorflow, Pytorch) and low-level hardware design in Verilog/VHDL creates a barrier to widespread adoption of FPGAs, which can be overcome with the help of High-Level Synthesis. hls4ml is a framework that translates Deep Neural Networks into annotated C…
Spiking Neural Networks (SNN) are an emerging computation model, which uses event-driven activation and bio-inspired learning algorithms. SNN-based machine-learning programs are typically executed on tile- based neuromorphic hardware platforms, where each tile consists of a computation unit called crossbar, which maps neurons and synapses of the program. However, synthesizing such programs on an off-the-shelf neuromorphic hardware is challenging. This is because of the inherent resource and latency limitations of the hardware, which impact both model performance, e.g., accuracy, and hardware p…
