Hardware design languages achieve similar results on fixed microarchitecture
Quantifying the Effect of HCLs on a Fixed-Microarchitecture MXFP4 Accelerator
Hardware Architecture
Summary
Hardware construction languages (HCLs) help engineers build circuits more productively without changing the circuit design itself. The authors compared several common HCLs by implementing the exact same hardware design and checking how well they performed on the same chip. They found that all the different languages produced circuits that matched or even improved on hand-written designs in terms of size and speed. Differences in performance mainly came from how the languages handle arithmetic operations, not from the designs. This means choosing an HCL depends more on the tools and user preferences than on performance.
What this means in practice
- •For hardware design engineers: Select a hardware construction language based on ecosystem and interface needs rather than performance differences for fixed designs.
- •For embedded system developers: Use any of the evaluated hardware languages to implement quantization primitives in AI edge devices with predictable area and timing results.
Authors
Daniele Passaretti, Sajjad Tamimi, Nicola Dall'Ora
Abstract
Hardware Construction Languages (HCLs) aim to improve hardware design productivity while generating register-transfer-level (RTL) circuits without changing the designer's microarchitecture. However, most comparisons between HCLs are either qualitative or evaluate quality of results (QoR) across different designs, making it difficult to separate language effects from design effects. This paper compares the most widely used HCLs using the same fixed design, the OCP MXFP4 block dot product, a quantization primitive at the heart of edge Physical-AI inference, implemented as a single 12-stage, II=1 pipeline. A SystemVerilog baseline is followed by implementations in Chisel, SpinalHDL, Amaranth, Clash, Bluespec, and C++ for high-level synthesis (HLS). Every variant goes through the same flow on the same Artix-7 device set at 100 MhZ, driven by a RISC-V soft core. With the micro-architecture held constant, the comparison is clean: every variant meets timing, and the HCLs match or even undercut hand-written RTL in area. The remaining differences stem not from the algorithm but from how each back end lowers arithmetic, and from a single width choice that silently toggles DSP inference. Unlike HLS, where design decisions are limited to pragmas, the HCLs achieve comparable area and timing. Therefore, the choice comes down to ecosystem fit and interface needs rather than QoR.