DE-HNN: An effective neural model for Circuit Netlist representation
Authors:
Zhishang Luo,
Truong Son Hy,
Puoya Tabaghi,
Donghyeon Koh,
Michael Defferrard,
Elahe Rezaei,
Ryan Carey,
Rhett Davis,
Rajeev Jain,
Yusu Wang
Abstract:
The run-time for optimization tools used in chip design has grown with the complexity of designs to the point where it can take several days to go through one design cycle which has become a bottleneck. Designers want fast tools that can quickly give feedback on a design. Using the input and output data of the tools from past designs, one can attempt to build a machine learning model that predicts…
▽ More
The run-time for optimization tools used in chip design has grown with the complexity of designs to the point where it can take several days to go through one design cycle which has become a bottleneck. Designers want fast tools that can quickly give feedback on a design. Using the input and output data of the tools from past designs, one can attempt to build a machine learning model that predicts the outcome of a design in significantly shorter time than running the tool. The accuracy of such models is affected by the representation of the design data, which is usually a netlist that describes the elements of the digital circuit and how they are connected. Graph representations for the netlist together with graph neural networks have been investigated for such models. However, the characteristics of netlists pose several challenges for existing graph learning frameworks, due to the large number of nodes and the importance of long-range interactions between nodes. To address these challenges, we represent the netlist as a directed hypergraph and propose a Directional Equivariant Hypergraph Neural Network (DE-HNN) for the effective learning of (directed) hypergraphs. Theoretically, we show that our DE-HNN can universally approximate any node or hyperedge based function that satisfies certain permutation equivariant and invariant properties natural for directed hypergraphs. We compare the proposed DE-HNN with several State-of-the-art (SOTA) machine learning models for (hyper)graphs and netlists, and show that the DE-HNN significantly outperforms them in predicting the outcome of optimized place-and-route tools directly from the input netlists. Our source code and the netlists data used are publicly available at https://github.com/YusuLab/chips.git
△ Less
Submitted 16 April, 2024; v1 submitted 30 March, 2024;
originally announced April 2024.
Multi-agent Learning for Cooperative Large-scale Caching Networks
Authors:
Elahe Rezaei,
Hafez Eslami Manoochehri,
Babak Hossein Khalaj
Abstract:
Caching networks are designed to reduce traffic load at backhaul links, by serving demands from edge-nodes. In the past decades, many studies have been done to address the caching problem. However, in practice, finding an optimal caching policy is still challenging due to dynamicity of traffic and scalability caused by complex impact of caching strategy chosen by each individual cache on other par…
▽ More
Caching networks are designed to reduce traffic load at backhaul links, by serving demands from edge-nodes. In the past decades, many studies have been done to address the caching problem. However, in practice, finding an optimal caching policy is still challenging due to dynamicity of traffic and scalability caused by complex impact of caching strategy chosen by each individual cache on other parts of network. In this paper, we focus on cache placement to optimize the performance metrics such as hit ratio in cooperative large-scale caching networks. Our proposed solution, cooperative multi-agent based cache placement (CoM-Cache) is based on multi-agent reinforcement learning framework and can seamlessly track the content popularity dynamics in an on-line fashion. CoM-Cache is enable to solve the problems over a spectrum from isolated to interconnected caches and is designed flexibly to fit any caching networks. To deal with dimensionality issue, CoM-Cache exploits the property of locality of interactions among caches. The experimental results report CoM-Cache outperforms base-line schemes, however at the expense of reasonable additional complexity.
△ Less
Submitted 30 June, 2018;
originally announced July 2018.
