Cygnus: A Heterogeneous Octa-Core RISC-V Vector Processor for Digital Signal Processing Applications
IEEE Journal of Solid-State Circuits · 2026-08 · 61 (8) · 4503-4517
DOI: 10.1109/jssc.2026.3676416 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Journal of Solid-State Circuits, 2026-08. Verfasst von Vikram Jain und 11 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Gefördert von: NSF CISE Community Research Infrastructure, NSF Pathways to Enable Open-Source Ecosystems.
Abstract (Originalsprache)
We introduce Cygnus, an energy-efficient octa-core RISC-V vector processor, compliant with the RISC-V vector extension (RVV) 1.0 specification, specifically designed for digital signal processing (DSP) and robotics applications. This system-on-chip (SoC) features dynamic instruction scheduling with short-vector lengths at the vector core level and employs a big/little architecture with tightly coupled memory (TCM) at the SoC level. The TCM design enables efficient data streaming across the cores, bypassing the cache hierarchy for streaming applications. Cygnus achieves a high average utilization of 90% across general matrix multiply (GEMM) and convolution (CONV) kernels, demonstrating energy efficiencies of 414 GOPS/W for INT8 GEMM and 109 GFLOPS/W for FP32 GEMM. In a typical computer vision kernel that implements denoising using a non-local means (NLMs) algorithm, the SoC provides a latency of 3.7 ms with an energy consumption of $518~\mu $ J per frame. Moreover, we demonstrate the adaptability of our approach using our SoC in a drone control application that employs a model-predictive control (MPC) algorithm based on TinyMPC. Cygnus, running the Zephyr real-time operating system (RTOS), manages control for three drones while optimizing performance through vectorization, achieving over $3.5\times $ improvement in solve time and significant reductions in position-tracking error at higher clock frequencies. The implementation demonstrates enhanced the real-time performance and accuracy compared with baseline scalar methods.
Förderung
- NSF CISE Community Research Infrastructure
- NSF Pathways to Enable Open-Source Ecosystems
Zitieren
APA
Jain, V., Grubb, D., Zhao, J., Anderson, K., Ho, K., Nikiforov, D., Dong, S. K., Chi, Y., Schwarz, E., Asanovic, K., Shao, Y. S., & Nikolic, B. (2026). Cygnus: A Heterogeneous Octa-Core RISC-V Vector Processor for Digital Signal Processing Applications. IEEE Journal of Solid-State Circuits, 61(8), 4503-4517. https://doi.org/10.1109/jssc.2026.3676416
Vancouver
Jain V, Grubb D, Zhao J, Anderson K, Ho K, Nikiforov D, et al. Cygnus: A Heterogeneous Octa-Core RISC-V Vector Processor for Digital Signal Processing Applications. IEEE Journal of Solid-State Circuits. 2026;61(8):4503-4517. doi:10.1109/jssc.2026.3676416
BibTeX
@article{Jain2026,
author = {Jain, Vikram and Grubb, Daniel and Zhao, Jerry and Anderson, Kevin and Ho, Ken and Nikiforov, Dima and Dong, Shengjun Kris and Chi, Yufeng and Schwarz, Ella and Asanovic, Krste and Shao, Yakun Sophia and Nikolic, Borivoje},
title = {Cygnus: A Heterogeneous Octa-Core RISC-V Vector Processor for Digital Signal Processing Applications},
journal = {IEEE Journal of Solid-State Circuits},
year = {2026},
volume = {61},
number = {8},
pages = {4503--4517},
doi = {10.1109/jssc.2026.3676416}
}
EUVIMED