The Memory Allocation problem for neural networks can be represented as a two-dimensional optimization problem. The neural network is allocated into limited memory space while allocating as much data as possible into the low latency memory. Our solution is based on a generalization of graph coloring, edge-to-node transformation and considers the order in which the graph nodes are colored. We observed improvement of more than 40% in SRAM memory bandwidth in various neural networks.
|Title of host publication||ICDCN 2022 - Proceedings of the 2022 International Conference on Distributed Computing and Networking|
|Publisher||Association for Computing Machinery|
|Number of pages||2|
|State||Published - 4 Jan 2022|
|Event||23rd International Conference on Distributed Computing and Networking, ICDCN 2022 - Virtual, Online, India|
Duration: 4 Jan 2022 → 7 Jan 2022
|Name||ACM International Conference Proceeding Series|
|Conference||23rd International Conference on Distributed Computing and Networking, ICDCN 2022|
|Period||4/01/22 → 7/01/22|
Bibliographical notePublisher Copyright:
© 2022 Owner/Author.
- Graph Coloring
- Memory Management
- Neural Networks