High-level Synthesis Design Research Articles

CollectiveHLS: A Collaborative Approach to High-Level Synthesis Design Optimization

High-Level Synthesis (HLS) has played a pivotal role in making FPGAs accessible to a broader audience by facilitating high-level device programming and rapid microarchitecture customization through the use of directives. However, manually selecting the right directives can be a formidable challenge for programmers lacking a hardware background. This paper presents CollectiveHLS, an ultra-fast, knowledge-driven approach to optimizing HLS designs. It automates the identification and application of optimal directive configurations from the original source code, focusing on minimizing design latency and ensuring synthesizability. This optimization approach is entirely data-driven, offering a generalized HLS tuning solution without reliance on Quality of Result (QoR) models or meta-heuristics. CollectiveHLS is designed, implemented, and evaluated using around 60 applications sourced from well-established benchmark suites and GitHub repositories, all running on a Xilinx UltraScale + MPSoC ZCU104. It achieves an average geometric mean speedup of up to \(23.1\times\) compared to the official source code without directives, while maintaining synthesizability and feasibility rates of 100% and 96.6%, respectively, matching those of Vitis, the industry-standard framework for FPGA acceleration. Comparisons with resource over-provisioning, traditional genetic algorithm-based Design Space Exploration (DSE), and State-of-the-Art (SotA) approaches demonstrate that CollectiveHLS produces designs of comparable quality \(14.6\times\) faster on average. These results underscore the potential of our approach as an ultra-fast and automated solution for HLS optimization.

Decomposition based estimation of distribution algorithm for high-level synthesis design space exploration

High-Level Synthesis (HLS) has evolved significantly due to the increasing complexity of integrated circuit design and the demand for efficient design methodologies. HLS, which raises the abstraction level of design specification, allows designers to focus on hardware functionality, thus enhancing productivity and reducing verification efforts. However, a key challenge in HLS is efficiently exploring the vast design space to find the Pareto-optimal designs. In this paper, we introduce a novel approach for multi-objective design space exploration in HLS. Our methodology decomposes the design space exploration problem into simpler sub-problems using the Multi-Objective Evolutionary Algorithm based on Decomposition (MOEA/D) framework and utilizes the Estimation of Distribution Algorithm (EDA) to build a probabilistic model for generating candidate solutions, thereby reducing the required number of expensive synthesis runs. Experimental results show that the proposed method has a faster convergence speed and reduces the number of syntheses by 24.34% to 32.01%, which significantly outperforms state-of-the-art works. Our approach achieves superior Pareto fronts with the lowest average ADRS value, outperforming Lattice-expl, ϵ -Constraint GA, and NSGA-II by 85.64%, 39.90%, and 33.31% respectively.

PASTA: Programming and Automation Support for Scalable Task-Parallel HLS Programs on Modern Multi-Die FPGAs

In recent years, the adoption of FPGAs in datacenters has increased, with a growing number of users choosing High-Level Synthesis (HLS) as their preferred programming method. While HLS simplifies FPGA programming, one notable challenge arises when scaling up designs for modern datacenter FPGAs that comprise multiple dies. The extra delays introduced due to die crossings and routing congestion can significantly degrade the frequency of large designs on these FPGA boards. Due to the gap between HLS design and physical design, it is challenging for HLS programmers to analyze and identify the root causes, and fix their HLS design to achieve better timing closure. Recent efforts have aimed to address these issues by employing coarse-grained floorplanning and pipelining strategies on task-parallel HLS designs where multiple tasks run concurrently and communicate through FIFO stream channels. However, many applications are not streaming friendly and many existing accelerator designs heavily rely on buffer channel based communication between tasks. In this work, we take a step further to support a task-parallel programming model where tasks can communicate via both FIFO stream channels and buffer channels. To achieve this goal, we design and implement the PASTA framework, which takes a large task-parallel HLS design as input and automatically generates a high-frequency FPGA accelerator via HLS and physical design co-optimization. Our framework introduces a latency-insensitive buffer channel design, which supports memory partitioning and ping-pong buffering while remaining compatible with vendor HLS tools. On the frontend, we provide an easy-to-use programming model for utilizing the proposed buffer channel; while on the backend, we implement efficient placement and pipelining strategies for the proposed buffer channel. To validate the effectiveness of our framework, we test it on four widely used Rodinia HLS benchmarks and two real-world accelerator designs and show an average frequency improvement of 25%, with peak improvements of up to 89% on AMD/Xilinx Alveo U280 boards compared to Vitis HLS baselines.

FADO: Floorplan-Aware Directive Optimization Based on Synthesis and Analytical Models for High-Level Synthesis Designs on Multi-Die FPGAs

Multi-die FPGAs are widely adopted for large-scale accelerators, but optimizing high-level synthesis designs on these FPGAs faces two challenges. First, the delay caused by die-crossing nets creates an NP-hard floorplanning problem. Second, traditional directive optimization cannot consider resource constraints on each die or the timing issue incurred by the die-crossings. Furthermore, the high algorithmic complexity and the large scale lead to extended runtime for legalizing the floorplan of HLS designs under different directive configurations. To co-optimize the directives and floorplan of HLS designs on multi-die FPGAs, we formulate the co-search based on bin-packing variants and present two iterative optimization flows. The first (FADO 1.0) relies on a pre-built QoR library. It involves a greedy, latency-bottleneck-guided directive search, and an incremental floorplan legalization. Compared with a global floorplanning solution, it takes 693X~4925X shorter search time and achieves 1.16X~8.78X better design performance, measured in workload execution time. To remove the time-consuming QoR library generation, the second flow (FADO 2.0) integrates an analytical QoR model and redesigns the directive search to accelerate convergence. Through experiments on mixed dataflow and non-dataflow designs, compared with 1.0, FADO 2.0 further yields a 1.40X better design performance on average after implementation on the Alveo U250 FPGA.

Fast and Inexpensive High-Level Synthesis Design Space Exploration: Machine Learning to the Rescue

High-Level Synthesis (HLS) has multiple significant advantages over traditional RT-level design flows. One in particular that we address in this work is the ability to generate multiple functional equivalent design variants with unique trade-offs like area, performance and power from the same behavioral description. This is typically done by setting synthesis options in the form or pragmas (comments) to mainly control how to synthesize arrays (RAM or registers), loops (unroll, partially unroll, no unroll or pipeline) and functions (inline or not). Setting different pragma combinations lead to these different design implementations. Out of all the pragma combinations the designer is typically only interested in those that lead to the Pareto-optimal designs. Fortunately, this search can be automated, but unfortunately, the search space to find these pragma combinations grows supra-linearly with the number of pragma settings. Thus, fast and efficient heuristics are needed. These heuristics generate a new pragma combination and then evaluate their effect by synthesizing (HLS) it. The most time-consuming part of this process is having to execute a full synthesis (HLS) on the behavioral description for every new pragma combination. One obvious way to accelerate the exploration is to parallelize the exploration process using a multithreaded heuristic. The theoretical speedup should match the number of parallel threads. The main problem with this approach is that every HLS invokation requires to check out a HLS tool license. This license is not released until the synthesis process has finished. This implies that the maximum number of parallel threads is restricted by the number of available licenses, which in the ASIC case are extremely expensive. On the contrary, FPGA vendors make their HLS tools free. Thus, it is tempting to investigate if FPGA HLS tools can be used to find the Pareto-optimal designs in the ASIC case. To address this, in this work we present a dedicated multithreaded parallel HLS design space explorer (DSE) based on transfer learning that is able to accelerate HLS DSE for ASICs by targeting first FPGAs and using machine learning to convert the exploration results obtained to find the optimal ASIC equivalent. Experimental results show the effectiveness and robustness of our approach.

HLS‐based swarm intelligence driven optimized hardware IP core for linear regression‐based machine learning

AbstractLinear Regression (LR), as one of the essential Machine Learning (ML) models, incurs massive data crunching during the training phase based on many data points. Considering the computationally intensive nature in the LR models, an optimized dedicated hardware IP core design can be very effective. This paper proposes the following novelties: (a) an optimized hardware IP core design of linear regression‐based machine learning model using high‐level synthesis (HLS). More specifically, independent application specific datapath architectures of hardware IP for computing optimal bias and intercepts and cost function in LR‐ML are presented here; (b) an optimized hardware IP core design of LR based ML model by deducing dependency graph from its corresponding mathematical foundation; (c) register transfer level (RTL) design, using HLS, of the optimized LR based ML hardware IP core for computing cost function; (d) linear regression IP core design using multi‐layered tree‐height transformation (THT) and swarm intelligence based architectural exploration for optimized HLS design.

FPGA implementation of QUasi-Affine TRansformation evolutionary algorithm

The QUasi-Affine TRansformation Evolutionary (QUATRE) algorithm, a new intelligence optimization algorithm, has been widely used in many optimization fields. In this paper, a hardware-based QUATRE algorithm is designed and implemented on a field-programmable gate array (FPGA). To facilitate the implementation of the QUATRE algorithm on hardware, this paper simplifies the co-evolutionary matrix generation process. Compared with the original QUATRE algorithm, the simplified QUATRE algorithm may reduce latency and resource occupation. The Vivado High-Level Synthesis (HLS) design tool is used to complete the IP core design of the QUATRE algorithm. Through the benchmark function test under different population sizes, compared with the QUATRE algorithm implemented by software, both the running speed and optimization performance of the QUATRE algorithm implemented by hardware are significantly better than the former. Compared with the GA, DE, and PSO algorithms implemented by hardware, the QUATRE algorithm also shows strong competitiveness.

AutoScaleDSE: A Scalable Design Space Exploration Engine for High-Level Synthesis

High-Level Synthesis (HLS) has enabled users to rapidly develop designs targeted for FPGAs from the behavioral description of the design. However, to synthesize an optimal design capable of taking better advantage of the target FPGA, a considerable amount of effort is needed to transform the initial behavioral description into a form that can capture the desired level of parallelism. Thus, a design space exploration (DSE) engine capable of optimizing large complex designs is needed to achieve this goal. We present a new DSE engine capable of considering code transformation, compiler directives (pragmas), and the compatibility of these optimizations. To accomplish this, we initially express the structure of the input code as a graph to guide the exploration process. To appropriately transform the code, we take advantage of ScaleHLS based on the multi-level compiler infrastructure (MLIR). Finally, we identify problems that limit the scalability of existing DSEs, which we name the “design space merging problem.” We address this issue by employing a Random Forest classifier that can successfully decrease the number of invalid design points without invoking the HLS compiler as a validation tool. We evaluated our DSE engine against the ScaleHLS DSE, outperforming it by a maximum of 59×. We additionally demonstrate the scalability of our design by applying our DSE to large-scale HLS designs, achieving a maximum speedup of 12× for the benchmarks in the MachSuite and Rodinia set.

High-Level Synthesis Design of Scalable Ultrafast Ultrasound Beamformer With Single FPGA.

Ultrafast ultrasound imaging is essential for advanced ultrasound imaging techniques such as ultrasound localization microscopy (ULM) and functional ultrasound (fUS). Current ultrafast ultrasound imaging is challenged by the ultrahigh data bandwidth associated with the radio frequency (RF) signal, and by the latency of the computationally expensive beamforming process. As such, continuous ultrafast data acquisition and beamforming remain elusive with existing software beamformers based on CPUs or GPUs. To address these challenges, the proposed work introduces a novel method of implementing an ultrafast ultrasound beamformer specifically for ultrafast plane wave imaging (PWI) on a field programmable gate array (FPGA) by using high-level synthesis. A parallelized implementation of the beamformer on a single FPGA was proposed by 1) utilizing a delay compression technique to reduce the delay profile size, which enables both run-time pre-calculated delay profile loading from external memory and delay reuse, 2) vectorizing channel data fetching which is enabled by delay reuse, and 3) using fixed summing networks to reduce consumption of logic resources. Our proposed method presents two unique advantages over current FPGA beamformers: 1) high scalability that allows fast adaptation to different FPGA resources and beamforming speed demands by using Xilinx High-Level Synthesis as the development tool, and 2) allow a compact form factor design by using a single FPGA to complete the beamforming instead of multiple FPGAs. Current Xilinx FPGAs provide the capabilities of connecting up to 1024 ultrasound channels with a single FPGA and the newest JESD204B interface analog front end (AFE). This channel count is much more than the channel count needed by current linear arrays, which normally have 128 or 256 channels. With the proposed method, a sustainable average beamforming rate of 4.83 G samples/second in terms of input raw RF sample was achieved. The resulting image quality of the proposed beamformer was compared with the software beamformer on the Verasonics Vantage system for both phantom imaging and in vivo imaging of a mouse brain. Multiple imaging schemes including B-mode, power Doppler and ULM were assessed to verify that the image quality was not compromised for speed.

A Parameterized Parallel Design Approach to Efficient Mapping of CNNs onto FPGA

In recent years, Convolution Neural Networks (CNNs) have been widely applied to some artificial intelligence (AI) systems such as computer vision. Among many existing hardware accelerators, FPGA is regarded as a suitable platform for the implementation of CNNs because of its high energy efficiency and flexible reconfigurability. In this paper, a parameterized design approach is proposed to explore the maximum parallelism that could be possibly implemented in mapping a CNN algorithm onto targeted FPGA resources. Four types of parallelism are employed in our parameterized design to fully exploit the processing resources available in FPGA. Meanwhile, a hardware library consisting of a set of modules is established to accommodate various CNN models. Further, an algorithm is proposed to find the optimal level of parallelism dedicated to a constrained amount of resources. As a case study, the typical LeNet-5 is implemented on Xilinx Zynq7020. Compared with the existing works using the high-level synthesis design flow, our design obtains higher FPS and lower latency under the premise of using fewer LUTs and FFs.

Exploring Sparse Visual Odometry Acceleration With High-Level Synthesis

Visual Odometry (VO) systems are widely used to determine the position and orientation of a robot or camera in an unknown environment. They are deployed on resource-constrained platforms, such as drones and Virtual Reality (VR) or Augmented Reality (AR) headsets. VO systems harnessing modern System-on-Chip (SoCs) with integrated Field Programmable Gate Array (FPGA) have the potential to improve the overall systems performance. This paper explores the FPGA acceleration of sparse VO kernels using High-level Synthesis (HLS) as this kind of VO system has been designed to use with low-power SoCs. We show that both computational and data transfer overheads between the processing cores of the CPU of the SoC and the accelerators on the FPGA need to be optimized to obtain better end-to-end performance. This is a result of the additional data movement incurred when using an FPGA accelerator and also because of the sparse computational nature with predictable or random memory access patterns of the kernels involved. However, state-of-the-art HLS tools are not yet able to perform the required optimizations automatically because they usually assume that the kernels to be accelerated have dense computational patterns with regular memory access. In this paper we propose three, potentially generic, methods to reduce the data transfer between the CPU and the customised hardware kernels on the FPGA; these methods are: (a) approximation based on domain-specific knowledge, (b) image compression, and (c) the use of on-the-fly computation. We present a case study of the use of these methods on SVO, a state-of-the-art sparse VO system with a semi-direct front-end. We demonstrate that our proposed methods can reduce data transfer overhead to achieve better end-to-end performance and that they can be applied not only when using standard Xilinx HLS tools but also with other state-of-the-art HLS tools, such as HeteroFlow. Compared to the baseline performance of the original SVO software on an Arm CPU, our proposed methods assist the HLS and HeteroFlow designs to achieve a speedup of 2.4x and 2.14x, respectively, without noticeable accuracy loss. The HLS and HeteroFlow designs also achieve a 1.85x and 1.89x, respectively, improvement in energy efficiency on the SoC system used. Compared to the SVO software baseline running on the Intel Xeon CPU, our proposed methods assist the HLS and HeteroFlow designs to achieve 8.2x and 8.3x improvement in energy efficiency, respectively.

Graph Neural Networks for High-Level Synthesis Design Space Exploration

High-level Synthesis (HLS) Design-Space Exploration (DSE) aims at identifying Pareto-optimal synthesis configurations whose exhaustive search is unfeasible due to the design-space dimensionality and the prohibitive computational cost of the synthesis process. Within this framework, we address the design automation problem by proposing graph neural networks that jointly predict acceleration performance and hardware costs of a synthesized behavioral specification given optimization directives. Learned models can be used to rapidly approach the Pareto curve by guiding the DSE, taking into account performance and cost estimates. The proposed method outperforms traditional HLS-driven DSE approaches, by accounting for the arbitrary length of computer programs and the invariant properties of the input. We propose a novel hybrid control and dataflow graph representation that enables training the graph neural network on specifications of different hardware accelerators. Our approach achieves prediction accuracy comparable with that of state-of-the-art simulators without having access to analytical models of the HLS compiler. Finally, the learned representation can be exploited for DSE in unexplored configuration spaces by fine-tuning on a small number of samples from the new target domain. The outcome of the empirical evaluation of this transfer learning shows strong results against state-of-the-art baselines in relevant benchmarks.

Machine learning based fast and accurate High Level Synthesis design space exploration: From graph to synthesis

In this paper, we present a machine learning based High-Level Synthesis (HLS) design space explorer (DSE) that significantly reduces the exploration runtime while leading to very accurate results. In order to do so, we leverage the power of low level virtual machine (LLVM) to generate the features used in the machine learning (ML) model. The proposed design space explorer uses a modified version of simulated annealing (SA) algorithm, where initially the search space is sampled to generate the predictive model. In this work we used gradient boost regression algorithm as our preferred ML model and achieve comparable results as a full DSE that performs logic synthesis for each newly generated design.

PathDriver+: Enhanced Path-Driven Architecture Design for Flow-Based Microfluidic Biochips

Continuous-flow microfluidic biochips have attracted high research interest over the past years. Inside such a chip, fluid samples of milliliter volumes are efficiently transported between devices (e.g., mixers, heaters, etc.) to automatically perform various laboratory procedures in biology and biochemistry. Each transportation task, however, requires an exclusive flow path composed of multiple contiguous microchannels during its execution period. Excess/waste fluids, in the meantime, should be discarded by independent flow paths connected to waste ports. All these paths are etched in a very tiny chip area using multilayer soft lithography and driven by flow ports connecting with external pressure sources, forming a highly integrated chip architecture that determines the final performance of biochips. In this article, we propose a new and practical design flow called PathDriver+ (PD+) for the architecture design of microfluidic biochips, integrating the actual fluid manipulations into both high-level synthesis and physical design, which has never been considered in prior work. With this design flow, highly efficient chip architectures with a flow-path network that enables the actual fluid transportation and removal can be constructed automatically. Meanwhile, fluid volume management between devices and flow-path minimization are realized for the first time, thus, ensuring the correctness of assay outcomes while reducing the complexity of chip architectures. Additionally, diagonal channel routing is implemented to fundamentally improve the chip performance. The tradeoff between the numbers of channel intersections and fluidic ports is evaluated to further reduce the fabrication cost of biochips. The experimental results on multiple benchmarks confirm that the proposed design flow leads to high assay execution efficiency and low overall chip cost.

FastSim: A Fast Simulation Framework for High-Level Synthesis

High-level synthesis (HLS) is a well-established framework used to translate high-level algorithmic behaviors into hardware designs. Despite the enduring research efforts, a major prevailing bottleneck of HLS is the large gap between the design and verification processes. Presently, register transfer level (RTL) simulation is the primary platform used for HLS design verification. Although most of the state-of-the-art RTL simulators provide an abstracted user-friendly platform for verification, they are undesirably slow and sometimes incomprehensible to non-field-programmable gate array experts to debug. The alternative software simulators (C simulation) introduced by commercial HLS tools render faster simulation, but are not cycle accurate and are not capable to estimate design performance. In this article, we introduce an automatic cycle-accurate simulation tool, FastSim, that manipulates certain unique features of HLS design to extract a concise, well-indented, and debug-friendly C behavior from the synthesized RTL. Our simulation tool ensures RTL correctness, provides cycle accuracy, accurate performance estimation, and renders on an average around 300 times faster simulation compared to RTL simulators and comparable performance to that of software C simulators.

Learning from the Past: Efficient High-level Synthesis Design Space Exploration for FPGAs

The quest to democratize the use of Field-Programmable Gate Arrays (FPGAs) has given High-Level Synthesis (HLS) the final push to be widely accepted with FPGA vendors strongly supporting this VLSI design methodology to expand the FPGA user base. HLS takes as input an untimed behavioral description and generates efficient RTL (Verilog or VHDL). One major advantage of HLS is that it allows us to generate a variety of different micro-architectures from the same behavioral description by simply specifying different combination of synthesis options. In particular, commercial HLS tools make extensive use of synthesize directives in the form pragmas. This strength is also a weakness as it forces HLS users to fully understand how these synthesis options work and how they interact to efficiently set them to get a hardware implementation with the desired characteristics. Luckily, this process can be automated. Unfortunately, the search space grows supra-linearly with the number of synthesis options. To address this, this work proposes an automatic synthesis option tuner dedicated for FPGAs. We have explored a larger number of behavioral descriptions targeting ASICs and FPGAs and found out that due to the internal structure of the FPGA a large number of synthesis options combinations never lead to a Pareto-optimal design and, hence, the search space can be drastically reduced. Moreover, we make use of large database of DSE results that we have generated since we started working in this field to further accelerate the exploration process. For this, we use a technique based on perceptual hashing that allows our proposed explorer to recognize similar program structures in the new description to be explored and match them with structures in our database. This allows us to directly retrieve the pragma settings that lead to Pareto-optimal configurations. Experimental results show that the search space can be accelerated substantially while leading to finding most of the Pareto-optimal designs.

Unified FPGA Design for the HEVC Dequantization and Inverse Transform Modules

As the newest standard, the High Efficiency Video Coding (HEVC) is specially designed to minimize the bitrate for video data transfer and to support High Definition (HD) and ULTRA HD video resolutions at the cost of increasing computational complexity relative to earlier standards like the H.264. Therefore, real-time video decoding with HEVC decoder becomes a challenging task. However, the Dequantization and Inverse Transform (DE/IT) are one of the computationally intensive modules in the HEVC decoder which are used to reconstruct the residual block. Thus, in this paper, a unified hardware architecture is proposed to implement the HEVC DE/IT module for all Transform Unit (TU) block size, including 4 × 4, 8 × 8, 16 × 16 and 32 × 32. This architecture is designed using the High-Level Synthesis (HLS) and the Low-Level Synthesis (LLS) methods in order to compare and determine the best method to implement in real-time the DE/IT module. In fact, the C/C++ programming language is used to generate an optimized hardware design for DE/IT module through the Xilinx Vivado HLS tool. On the other hand, the LLS hardware architecture is designed by the VHSIC Hardware Description language (VHDL) and using the pipeline technique to decrease the processing time. The experimental results on the Xilinx XC7Z020 FPGA show that the LLS design increases the throughput in term of frame rate by 80% relative to HLS design with a 4.4% increase in the number of Look-Up Tables (LUTs). Compared with existing related works in literature, the proposed architectures demonstrate significant advantages in hardware cost and performance improvement.

DB4HLS: A Database of High-Level Synthesis Design Space Explorations

High-level synthesis (HLS) frameworks allow to easily specify a large number of variants of the same hardware design by only acting on optimization directives. Nonetheless, the hardware synthesis of implementations for all possible combinations of directive values is impractical even for simple designs. Addressing this shortcoming, many HLS design space exploration (DSE) strategies have been proposed to devise directive settings leading to high-quality implementations while limiting the number of synthesis runs. All these works require considerable efforts to validate the proposed strategies and/or to build the knowledge base employed to tune abstract models, as both tasks mandate the syntheses of large collections of implementations. Currently, such data gathering is performed <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ad hoc</i> : 1) leading to a lack of standardization, hampering comparisons between DSE alternatives; and 2) posing a very high burden to researchers willing to develop novel DSE strategies. Against this backdrop, we here introduce DB4HLS, a database of exhaustive HLS explorations comprising more than 100 000 design points collected over four years equivalent of synthesis time. The open structure of DB4HLS allows the incremental integration of new DSEs, which can be easily defined with a dedicated domain-specific language. We think that of our database, available at <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><uri>https://www.db4hls.inf.usi.ch/</uri></monospace> , will be a valuable tool for the research community investigating automated strategies for the optimization of HLS-based hardware designs.

AutoBridge: Coupling Coarse-Grained Floorplanning and Pipelining for High-Frequency HLS Design on Multi-Die FPGAs.

Despite an increasing adoption of high-level synthesis (HLS) for its design productivity advantages, there remains a significant gap in the achievable frequency between an HLS design and a handcrafted RTL one. A key factor that limits the timing quality of the HLS outputs is the difficulty in accurately estimating the interconnect delay at the HLS level. This problem becomes even worse when large HLS designs are implemented on the latest multi-die FPGAs. To tackle this challenge, we propose AutoBridge, an automated framework that couples a coarse-grained floorplanning step with pipelining during HLS compilation. First, our approach provides HLS with a view on the global physical layout of the design, allowing HLS to more easily identify and pipeline the long wires, especially those crossing the die boundaries. Second, by exploiting the flexibility of HLS pipelining, the floorplanner is able to distribute the design logic across multiple dies on the FPGA device without degrading clock frequency. This prevents the placer from aggressively packing the logic on a single die which often results in local routing congestion that eventually degrades timing. Since pipelining may introduce additional latency, we further present analysis and algorithms to ensure the added latency will not compromise the overall throughput. AutoBridge can be integrated into the existing CAD toolflow for Xilinx FPGAs. In our experiments with a total of 43 design configurations, we improve the average frequency from 147 MHz to 297 MHz (a 102% improvement) with no loss of throughput and a negligible change in resource utilization. Notably, in 16 experiments we make the originally unroutable designs achieve 274 MHz on average. The tool is available at https://github.com/Licheng-Guo/AutoBridge.

Cluster-Based Heuristic for High Level Synthesis Design Space Exploration

High Level Synthesis (HLS) frameworks allow to describe hardware designs in a high-level language (C/C++), without burdening developers with the error-prone task of specifying their implementation in detail. The HLS process is usually controlled by user-specified directives, which influence the implementation area and latency. Nonetheless, the correlation between directives and performance is often difficult to foresee and to quantify. Addressing this challenge, we herein propose a heuristic that, by only exploring a subset of possible configurations for an HLS design, is able to retrieve a close approximation of its Pareto Frontier of non-dominated implementations. Our framework identifies regions of interest in the design space, and iteratively searches for new solutions within such regions, or in their combinations. Experimental evidence across multiple benchmarks showcases that our approach to HLS design space exploration reaches better Pareto approximations, and with less required synthesis runs, with respect to State of the Art alternatives.