• Cite Icon51
  • https://doi.org/10.1145/3296957.3173171Copy DOI Icon

In-Memory Data Parallel Processor

Show More
  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

Recent developments in Non-Volatile Memories (NVMs) have opened up a new horizon for in-memory computing. Despite the significant performance gain offered by computational NVMs, previous works have relied on manual mapping of specialized kernels to the memory arrays, making it infeasible to execute more general workloads. We combat this problem by proposing a programmable in-memory processor architecture and data-parallel programming framework. The efficiency of the proposed in-memory processor comes from two sources: massive parallelism and reduction in data movement. A compact instruction set provides generalized computation capabilities for the memory array. The proposed programming framework seeks to leverage the underlying parallelism in the hardware by merging the concepts of data-flow and vector processing. To facilitate in-memory programming, we develop a compilation framework that takes a TensorFlow input and generates code for our in-memory processor. Our results demonstrate 7.5x speedup over a multi-core CPU server for a set of applications from Parsec and 763x speedup over a server-class GPU for a set of Rodinia benchmarks.

Similar Papers
  • Conference Article
  • Citations7

In-memory Data Flow Processor

  • Sep 01, 2017
  • Daichi Fujiki +2
  • Conference Article
  • Citations120

In-Memory Data Parallel Processor

  • Mar 19, 2018
  • Daichi Fujiki +2
  • Conference Article
  • Citations1

FeFET versus DRAM based PIM Architectures: A Comparative Study

  • Oct 03, 2022
  • Chirag Sudarshan +4
  • Conference Article
  • Citations12

On Endurance of Processing in (Nonvolatile) Memory

  • Jun 17, 2023
  • Salonik Resch +7
  • Conference Article
  • Citations3

AnalyzeThat: A Programmable Shared-Memory System for an Array of Processing-In-Memory Devices

  • May 01, 2017
  • Sangkuen Lee +3
  • Research Article

A programmable shared-memory system for an array of processing-in-memory devices

  • Aug 30, 2018
  • Cluster Computing
  • Sangkuen Lee +3
  • Research Article
  • Citations2

PIMSAB: A Processing-In-Memory System with Spatially-Aware Communication and Bit-Serial-Aware Computation

  • Nov 20, 2024
  • ACM Transactions on Architecture and Code Optimization
  • Siyuan Ma +7
  • Conference Article
  • Citations4

Exploiting Parallelism for Convolutional Connections in Processing-In-Memory Architecture

  • Jun 18, 2017
  • Yi Wang +2
  • Conference Article
  • Citations470

Pinatubo

  • Jun 05, 2016
  • Shuangchen Li +5
  • Research Article

Efficient associative processing in FPGA

  • Jun 01, 2026
  • Microprocessors and Microsystems
  • Jonathas Silveira +2
  • Research Article
  • Citations107

MRIMA: An MRAM-Based In-Memory Accelerator

  • Apr 11, 2019
  • IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
  • Shaahin Angizi +3
  • Research Article
  • Citations8

A Weighted Current Summation Based Mixed Signal DRAM-PIM Architecture for Deep Neural Network Inference

  • Jun 01, 2022
  • IEEE Journal on Emerging and Selected Topics in Circuits and Systems
  • Chirag Sudarshan +7
  • Research Article
  • Citations10

Achieving the Performance of All-Bank In-DRAM PIM With Standard Memory Interface: Memory-Computation Decoupling

  • Jan 01, 2022
  • IEEE Access
  • Yoonah Paik +3
  • Dissertation
  • Citations4

Automatic data distribution for massively parallel processors

  • Apr 16, 1997
  • Jordi García Almiñana
  • Conference Article
  • Citations1

MAPIM: Mat Parallelism for High Performance Processing in Non-volatile Memory Architecture

  • Mar 01, 2019
  • Joonseop Sim +5
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.