A 19.4-nJ/Decision, 364-K Decisions/s, In-Memory Random Forest Multi-Class Inference Accelerator

Mingu Kang, Sujan K. Gonugondla, Sungmin Lim, Naresh R. Shanbhag

Research output: Contribution to journalArticlepeer-review


This paper presents an integrated circuit (IC) realization of a random forest (RF) machine learning classifier in a 65-nm CMOS. Algorithm, architecture, and circuits are co-optimized to achieve aggressive energy and delay benefits by taking advantage of the inherent error resiliency derived from the ensemble nature of an RF classifier. Deterministic sub-sampling (DSS) and regularized decision trees reduce interconnect complexity, and avoid irregular memory access patterns and computations, thereby reducing the energy-delay product (EDP). The prototype IC also employs low-swing analog in-memory computations embedded in a standard 6T SRAM to enable massively parallel tree node comparisons, thereby minimizing the memory fetches and reducing the EDP further. The 65-nm CMOS prototype IC achieves a 3.1 × and 2.2 × improved energy efficiency and throughput leading to 6.8 × lower EDP compared to a conventional digital system at the same accuracies of 94% and 97.5% for two tasks: 1) eight-class traffic sign recognition and 2) face detection, respectively.

Original languageEnglish (US)
Pages (from-to)2126-2135
Number of pages10
JournalIEEE Journal of Solid-State Circuits
Issue number7
StatePublished - Jul 2018
Externally publishedYes


  • Accelerator
  • analog processing
  • in-memory computing
  • machine learning (ML)
  • random forest (RF)

ASJC Scopus subject areas

  • Electrical and Electronic Engineering


Dive into the research topics of 'A 19.4-nJ/Decision, 364-K Decisions/s, In-Memory Random Forest Multi-Class Inference Accelerator'. Together they form a unique fingerprint.

Cite this