Performance Analysis of Digital Processing-in-Memory through a Case Study on Convolutional-Neural-Network Acceleration
For researchers and engineers in hardware acceleration, this work provides a quantitative comparison highlighting the limitations of digital PIM for CNN acceleration, which is incremental as it confirms known bottlenecks.
This paper analyzes digital processing-in-memory (PIM) architectures for CNN acceleration, comparing them with GPUs through theoretical investigation and benchmarks. Results show digital PIM achieves up to 10x energy efficiency improvement over GPUs for memory-bound operations, but performance is limited by data movement and bit-serial computation.
Processing-in-Memory (PIM) architectures are evolving to minimize data movement by leveraging the same physical devices for both memory and logic functionalities. While analog PIM harnesses crossbar arrays for efficient approximate matrix-vector multiplication, digital PIM architectures facilitate massively-parallel bitwise operations for more general workloads. Recent works have extended digital PIM towards the full-precision acceleration of convolutional neural networks (CNNs), yet a comprehensive comparison with GPUs remains a gap in the literature that may illuminate the limitations of digital PIM. This paper aims to fill this void by conducting a thorough examination of CNN acceleration through an updated quantitative comparison with GPUs. Our approach begins with a theoretical investigation into various PIM architectures, shedding light on their performance characteristics and constraints. Subsequently, through a series of benchmarks spanning memory-bound vectored arithmetic to CNN acceleration, we provide insights into digital PIM performance that may guide the acceleration of applications in the future.