# Efficiency and Scalability of Multi-Lane Capsule Networks (MLCN)

## TL;DR

This paper investigates the efficiency and scalability of Multi-lane Capsule Networks (MLCN), demonstrating that MLCN is twice as efficient as traditional CapsNet with parallel processing and proposing a load balancing heuristic to improve performance.

## Contribution

It systematically examines MLCN's efficiency and scalability, introduces a load balancing heuristic, and demonstrates significant performance improvements.

## Key findings

- MLCN achieves 2x efficiency over original CapsNet with model-parallelism.
- A simple greedy heuristic improves load balancing, nearly doubling speed.
- Parallel lanes in MLCN enable better scalability on multiple GPUs.

## Abstract

Some Deep Neural Networks (DNN) have what we call lanes, or they can be reorganized as such. Lanes are paths in the network which are data-independent and typically learn different features or add resilience to the network. Given their data-independence, lanes are amenable for parallel processing. The Multi-lane CapsNet (MLCN) is a proposed reorganization of the Capsule Network which is shown to achieve better accuracy while bringing highly-parallel lanes. However, the efficiency and scalability of MLCN had not been systematically examined. In this work, we study the MLCN network with multiple GPUs finding that it is 2x more efficient than the original CapsNet when using model-parallelism. Further, we present the load balancing problem of distributing heterogeneous lanes in homogeneous or heterogeneous accelerators and show that a simple greedy heuristic can be almost 50% faster than a naive random approach.