PyHLS: Intermediate Representation for Versatile High-Level Synthesis
Salvato in:
| Pubblicato in: | International Journal of Electronics and Telecommunications vol. 71, no. 4 (2025) |
|---|---|
| Pubblicazione: |
Polish Academy of Sciences
|
| Soggetti: | |
| Accesso online: | Citation/Abstract Full Text - PDF |
| Tags: |
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
MARC
| LEADER | 00000nab a2200000uu 4500 | ||
|---|---|---|---|
| 001 | 3273776711 | ||
| 003 | UK-CbPIL | ||
| 022 | |a 2081-8491 | ||
| 022 | |a 2300-1933 | ||
| 022 | |a 0035-9386 | ||
| 022 | |a 0867-6747 | ||
| 035 | |a 3273776711 | ||
| 045 | 2 | |b d20251001 |b d20251231 | |
| 084 | |a 190537 |2 nlm | ||
| 245 | 1 | |a PyHLS: Intermediate Representation for Versatile High-Level Synthesis | |
| 260 | |b Polish Academy of Sciences |c 2025 | ||
| 513 | |a Journal Article | ||
| 520 | 3 | |a High-Level Synthesis (HLS) has become an established methodology to accelerate the development of FPGAbased systems by allowing algorithms to be written in high-level languages (HLLs) such as C/C++ or Python. Yet, for real-time physics experiments—including fusion plasma diagnostics, highenergy physics (HEP) detectors, and rare-event astrophysical triggers—conventional HLS still falls short in three essential aspects: determinism, portability, and auditability. Pragmas embedded in HLL code blur the separation between algorithmic intent and implementation details, coupling scientific software to a particular device or compiler version. This is particularly problematic in long-lived scientific projects such as ITER or the Pierre Auger Observatory, where systems must remain functional and maintainable over decades [4]–[6]. To address these challenges, we propose an Intermediate Representation (IR)-centric HLS flow—PyHLS—that explicitly introduces an abstraction layer between algorithm and Register- Transfer Level (RTL) design. The IR centralizes all performancecritical aspects: timing contracts (initiation interval, latency, jitter), concurrency (loop unrolling, pipelining), memory layout (banking, tiling, port allocation), and resource binding (DSPs, BRAMs, AI tiles). In this model, the algorithm is expressed in clean, testable Python code [1], [2], while device-specific optimizations are described in a structured IR graph. This IR is then lowered into a reusable VHDL microinstruction library [3], which serves as a portable middle layer across devices. By versioning and auditing IR graphs and instruction streams, PyHLS ensures reproducibility and traceability—critical properties in scientific computing where results must be verifiable years after deployment. The methodology builds upon earlier work in Python-based high-level synthesis, parameterizable metamodels, and algorithmic synthesis with multi-level compilers [8], [9], [11]. It incorporates systematic design space exploration (DSE), allowing parameter sweeps over IR attributes and early feasibility checks. The flow is complemented by a cycle-accurate microinstruction emulator, which validates both functionality and timing contracts before vendor toolchains are invoked, reducing iteration time and catching infeasible designs early. We demonstrate the motivation and applicability of this approach in two demanding domains. First, in plasma diagnostics at JET/ITER, where spectrometer and data acquisition systems must combine high bandwidth with deterministic latency [6]. Second, in trigger pipelines for astroparticle physics, where artificial neural networks (ANNs) and fuzzy-logic algorithms have been implemented directly in FPGA logic to discriminate rare events from large backgrounds [4], [5]. These use-cases highlight the need for explicit IR-level contracts and modularity: the same high-level algorithm must be portable across device generations, yet adapted to exploit specialized hardware resources such as DSP slices, systolic AI engines, or high-bandwidth memories. The contribution of this work is therefore threefold: 1) We formalize the role of an explicit IR in HLS, decoupling algorithms from implementation decisions and introducing contract-driven determinism. 2) We present a reusable VHDL microinstruction library and emulator that stabilize implementation and provide auditable artifacts. 3) We show how PyHLS extends naturally to heterogeneous FPGAs, mapping operators to emerging AI/ML blocks while maintaining scientific reproducibility and portability across decades. By unifying algorithmic specification, IR-based parameterization, and reusable microinstructions, PyHLS establishes a sustainable methodology for real-time physics experiments and beyond. In short: write the science once, retarget the hardware many times.High-Level Synthesis (HLS) has become an established methodology to accelerate the development of FPGAbased systems by allowing algorithms to be written in high-level languages (HLLs) such as C/C++ or Python. Yet, for real-time physics experiments—including fusion plasma diagnostics, highenergy physics (HEP) detectors, and rare-event astrophysical triggers—conventional HLS still falls short in three essential aspects: determinism, portability, and auditability. Pragmas embedded in HLL code blur the separation between algorithmic intent and implementation details, coupling scientific software to a particular device or compiler version. This is particularly problematic in long-lived scientific projects such as ITER or the Pierre Auger Observatory, where systems must remain functional and maintainable over decades [4]–[6]. To address these challenges, we propose an Intermediate Representation (IR)-centric HLS flow—PyHLS—that explicitly introduces an abstraction layer between algorithm and Register- Transfer Level (RTL) design. The IR centralizes all performancecritical aspects: timing contracts (initiation interval, latency, jitter), concurrency (loop unrolling, pipelining), memory layout (banking, tiling, port allocation), and resource binding (DSPs, BRAMs, AI tiles). In this model, the algorithm is expressed in clean, testable Python code [1], [2], while device-specific optimizations are described in a structured IR graph. This IR is then lowered into a reusable VHDL microinstruction library [3], which serves as a portable middle layer across devices. By versioning and auditing IR graphs and instruction streams, PyHLS ensures reproducibility and traceability—critical properties in scientific computing where results must be verifiable years after deployment. The methodology builds upon earlier work in Python-based high-level synthesis, parameterizable metamodels, and algorith-incorporates systematic design space exploration (DSE), allowing parameter sweeps over IR attributes and early feasibility checks. The flow is complemented by a cycle-accurate microinstruction emulator, which validates both functionality and timing contracts before vendor toolchains are invoked, reducing iteration time and catching infeasible designs early. We demonstrate the motivation and applicability of this approach in two demanding domains. First, in plasma diagnostics at JET/ITER, where spectrometer and data acquisition systems must combine high bandwidth with deterministic latency [6]. Second, in trigger pipelines for astroparticle physics, where artificial neural networks (ANNs) and fuzzy-logic algorithms have been implemented directly in FPGA logic to discriminate rare events from large backgrounds [4], [5]. These use-cases highlight the need for explicit IR-level contracts and modularity: the same high-level algorithm must be portable across device generations, yet adapted to exploit specialized hardware resources such as DSP slices, systolic AI engines, or high-bandwidth memories. The contribution of this work is therefore threefold: 1) We formalize the role of an explicit IR in HLS, decoupling algorithms from implementation decisions and introducing contract-driven determinism. 2) We present a reusable VHDL microinstruction library and emulator that stabilize implementation and provide auditable artifacts. 3) We show how PyHLS extends naturally to heterogeneous FPGAs, mapping operators to emerging AI/ML blocks while maintaining scientific reproducibility and portability across decades. By unifying algorithmic specification, IR-based parameterization, and reusable microinstructions, PyHLS establishes a sustainable methodology for real-time physics experiments and beyond. In short: write the science once, retarget the hardware many times.H | |
| 653 | |a Software | ||
| 653 | |a Plasma | ||
| 653 | |a Computation | ||
| 653 | |a Artifacts | ||
| 653 | |a Data acquisition | ||
| 653 | |a Bandwidths | ||
| 653 | |a Banking | ||
| 653 | |a Artificial neural networks | ||
| 653 | |a Separation | ||
| 653 | |a Mapping | ||
| 653 | |a High level synthesis | ||
| 653 | |a High level languages | ||
| 653 | |a Modularity | ||
| 653 | |a Contracts | ||
| 653 | |a Python | ||
| 653 | |a Representations | ||
| 653 | |a Logic | ||
| 653 | |a Data acquisition systems | ||
| 653 | |a Coupling | ||
| 653 | |a Detectors | ||
| 653 | |a Space exploration | ||
| 653 | |a Parameterization | ||
| 653 | |a Physics | ||
| 653 | |a Methodology | ||
| 653 | |a Operators (mathematics) | ||
| 653 | |a Experiments | ||
| 653 | |a Pipelining (computers) | ||
| 653 | |a Observatories | ||
| 653 | |a Specifications | ||
| 653 | |a Neural networks | ||
| 653 | |a Algorithms | ||
| 653 | |a Compilers | ||
| 653 | |a Libraries | ||
| 653 | |a Determinism | ||
| 653 | |a Engines | ||
| 653 | |a Parameters | ||
| 653 | |a Reproducibility | ||
| 653 | |a Creeks & streams | ||
| 773 | 0 | |t International Journal of Electronics and Telecommunications |g vol. 71, no. 4 (2025) | |
| 786 | 0 | |d ProQuest |t Advanced Technologies & Aerospace Database | |
| 856 | 4 | 1 | |3 Citation/Abstract |u https://www.proquest.com/docview/3273776711/abstract/embedded/L8HZQI7Z43R0LA5T?source=fedsrch |
| 856 | 4 | 0 | |3 Full Text - PDF |u https://www.proquest.com/docview/3273776711/fulltextPDF/embedded/L8HZQI7Z43R0LA5T?source=fedsrch |