Marsellus: A Heterogeneous RISC-V AI-IoT End-Node SoC With 2–8 b DNN Acceleration and 30%-Boost Adaptive Body Biasing
Authors:
- Francesco ContiFCFrancesco Conti
- Gianna PaulinGPGianna Paulin
- Angelo GarofaloAGAngelo Garofalo
- Davide RossiDRDavide Rossi
- Alfio Di MauroAMAlfio Di Mauro
- Georg RutishauserGRGeorg Rutishauser
- Gianmarco OttaviGOGianmarco Ottavi
- Manuel EggimanMEManuel Eggiman
- Hayate OkuharaHOHayate Okuhara
- Luca BeniniLBLuca Benini
Top Cited Papers
- 3 October 2023
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 59 (1) , 128-142
Abstract
Emerging artificial intelligence-enabled Internet-of-Things (AI-IoT) system-on-chip (SoC) for augmented reality, personalized healthcare, and nanorobotics need to run many diverse tasks within a power envelope of a few tens of mW over a wide range of operating conditions: compute-intensive but strongly quantized deep neural network (DNN) inference, as well as signal processing and control requiring high-precision floating point. We present MARSELLUS, an all-digital heterogeneous SoC for AI-IoT end-nodes fabricated in GlobalFoundries 22 nm FDX that combines: 1) a general-purpose cluster of 16 RISC-V digital signal processing (DSP) cores attuned for the execution of a diverse range of workloads exploiting 4- and 2-bit arithmetic extensions ( XpulpNN ), combined with fused multiply accumulate (MAC) and LOAD operations and floating-point support; 2) a 2–8 bit reconfigurable binary engine (RBE) to accelerate $A3 3$ and $A1 1$ (pointwise) convolutions in DNNs; 3) a set of ON-chip monitoring (OCM) blocks connected to an adaptive body biasing (ABB) generator and a hardware control loop, enabling on- the-fly adaptation of transistor threshold voltages. MARSELLUS achieves up to 180 Gop/s or 3.32 Top/s/W on 2-bit precision arithmetic in software, and up to 637 Gop/s or 12.4 Top/s/W on hardware-accelerated DNN layers.Keywords
All Related Versions
Funding Information
- European Union (EU) Horizon 2020 Project WiPLASH (863337)
- EU Horizon Europe Project NeuroSoC (101070634)
- Key Digital Technologies (KDT) Joint Undertaking Project TRISTAN (101095947)
- Convolve Project evaluated by the EU Horizon Europe (101070374)
- Swiss State Secretariat for Education Research and Innovation (0000DONOTUSETHIS0000.00150)
This publication has 33 references indexed in Scilit:
- A Survey of Convolutional Neural Networks: Analysis, Applications, and ProspectsIEEE Transactions on Neural Networks and Learning Systems, 2021
- DORY: Automatic End-to-End Deployment of Real-World DNNs on Low-Cost IoT MCUsIEEE Transactions on Computers, 2021
- Efficient Transform Algorithms for Parallel Ultra-Low-Power IoT End NodesIEEE Embedded Systems Letters, 2021
- SleepRunner: A 28-nm FDSOI ULP Cortex-M0 MCU With ULL SRAM and UFBR PVT Compensation for 2.6–3.6-μW/DMIPS 40–80-MHz Active Mode and 131-nW/kB Fully Retentive Deep-Sleep ModeIEEE Journal of Solid-State Circuits, 2021
- 35.2 A 0.021mm2 PVT-Aware Digital-Flow-Compatible Adaptive Back-Biasing Regulator with Scalable Drivers Achieving 450% Frequency Boosting and 30% Power Reduction in 22nm FDSOI Technology2021
- HAWQ: Hessian AWare Quantization of Neural Networks With Mixed-Precision2019
- Data-Free Quantization Through Weight Equalization and Bias Correction2019
- XNOR Neural Engine: A Hardware Accelerator IP for 21.6-fJ/op Binary Neural Network InferenceIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2018
- Near-Threshold RISC-V Core With DSP Extensions for Scalable IoT Endpoint DevicesIEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2017
- A 60 GOPS/W, −1.8 V to 0.9 V body bias ULP cluster in 28 nm UTBB FD-SOI technologySolid-State Electronics, 2016