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.
ScaleHLS: a scalable high-level synthesis framework with multi-level transformations and optimizations: invited
This paper presents an enhanced version of a scalable HLS (High-Level Synthesis) framework named ScaleHLS, which can compile HLS C/C++ programs and PyTorch models to highly-efficient and synthesizable C++ designs. The original version of ScaleHLS achieved significant speedup on both C/C++ kernels and PyTorch models [14]. In this paper, we first highlight the key features of ScaleHLS on tackling the challenges present in the representation, optimization, and exploration of large-scale HLS designs. To further improve the scalability of ScaleHLS, we then propose an enhanced HLS transform and analysis library supported in both C++ and Python, and a new design space exploration algorithm to handle HLS designs with hierarchical structures more effectively. Comparing to the original ScaleHLS, our enhanced version improves the speedup by up to 60.9× on FPGAs. ScaleHLS is fully open-sourced at https://github.com/hanchenye/scalehls.
more »
« less
- Award ID(s):
- 2117997
- NSF-PAR ID:
- 10477705
- Publisher / Repository:
- ACM
- Date Published:
- Page Range / eLocation ID:
- 1355 to 1358
- Format(s):
- Medium: X
- Location:
- San Francisco California
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
-
null (Ed.)High-level synthesis (HLS) raises the level of design abstraction, expedites the process of hardware design, and enriches the set of final designs by automatically translating a behavioral specification into a hardware implementation. To obtain different implementations, HLS users can apply a variety of knobs, such as loop unrolling or function inlining, to particular code regions of the specification. The applied knob configuration significantly affects the synthesized design's performance and cost, e.g., application latency and area utilization. Hence, HLS users face the design-space exploration (DSE) problem, i.e. determine which knob configurations result in Pareto-optimal implementations in this multi-objective space. Whereas it can be costly in time and resources to run HLS flows with an enormous number of knob configurations, machine learning approaches can be employed to predict the performance and cost. Still, they require a sufficient number of sample HLS runs. To enhance the training performance and reduce the sample complexity, we propose a transfer learning approach that reuses the knowledge obtained from previously explored design spaces in exploring a new target design space. We develop a novel neural network model for mixed-sharing multi-domain transfer learning. Experimental results demonstrate that the proposed model outperforms both single-domain and hard-sharing models in predicting the performance and cost at early stages of HLS-driven DSE.more » « less
-
HeteroGen: transpiling C to heterogeneous HLS code with automated test generation and program repairDespite the trend of incorporating heterogeneity and specialization in hardware, the development of heterogeneous applications is limited to a handful of engineers with deep hardware expertise. We propose HeteroGen that takes C/C++ code as input and automatically generates an HLS version with test behavior preservation and better performance. Key to the success of HeteroGen is adapting the idea of search-based program repair to the heterogeneous computing domain, while addressing two technical challenges. First, the turn-around time of HLS compilation and simulation is much longer than the usual C/C++ compilation and execution time; therefore, HeteroGen applies pattern-oriented program edits guided by common fix patterns and their dependences. Second, behavior and performance checking requires testing, but test cases are often unavailable. Thus, HeteroGen auto-generates test inputs suitable for checking C to HLS-C conversion errors, while providing high branch coverage for the original C code. An evaluation of HeteroGen shows that it produces an HLS-compatible version for nine out of ten real-world heterogeneous applications fully automatically, applying up to 438 lines of edits to produce an HLS version 1.63x faster than the original version.more » « less
-
The design of heterogeneous systems that include domain specific accelerators is a challenging and time-consuming process. While taking into account area constraints, designers must decide which parts of an application to accelerate in hardware and which to leave in software. Moreover, applications in domains such as Extended Reality (XR) offer opportunities for various forms of parallel execution, including loop level, task level and pipeline parallelism. To assist the design process and expose every possible level of parallelism, we present Trireme , a fully automated tool-chain that explores multiple levels of parallelism and produces domain specific accelerator designs and configurations that maximize performance, given an area budget. FPGA SoCs were used as target platforms and Catapult HLS [7] was used to synthesize RTL using a commercial 12nm FinFET technology. Experiments on demanding benchmarks from the XR domain revealed a speedup of up to 20 ×, as well as a speedup of up to 37 × for smaller applications, compared to software-only implementations.more » « less
-
As the technology node of VLSI designs advances to sub10 nm, two interconnect-centric metrics of a circuit, the interconnect complexity (either number of interconnects or wirelength/WL) and congestion, become critically important across all design stages alongside conventional resource or function-unit (FU)-centric metrics like area/number-of-FUs and leakage power. High Level synthesis (HLS), one of the earliest and most impactful design stages, rarely monitors interconnect metrics, which makes their recovery at later stages very difficult. HLS algorithms and tools typically perform FU-centric minimization via operation scheduling, module selection (S&MS) and binding. As a consequence, it mostly overlooks interconnect-based metrics. In this paper, we explore whether this can adversely affect interconnect metrics, and in general explore the correlation between FU-centric optimization in S&MS, and the resulting interconnect metrics co-optimized (along with FU metrics) in the later binding stage(s). For this purpose we develop a probabilistic analysis for post-scheduling binding to estimate interconnect metrics, and verify its accuracy by comparison to empirical results across different scheduling techniques that generate different degrees of FU optimization. Based on both empirical and analytical results we predict how interconnects metrics will pan out with different degrees of FU optimization. Finally, based on our analysis, we also provide suggestions to improve interconnect metrics for whatever FU optimization degree an available S&MS technique can achieve.more » « less
