Summary of Rankmap: Priority-aware Multi-dnn Manager For Heterogeneous Embedded Devices, by Andreas Karatzas et al.
RankMap: Priority-Aware Multi-DNN Manager for Heterogeneous Embedded Devices
by Andreas Karatzas, Dimitrios Stamoulis, Iraklis Anagnostopoulos
First submitted to arxiv on: 26 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Distributed, Parallel, and Cluster Computing (cs.DC); Emerging Technologies (cs.ET)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A novel priority-aware manager called RankMap is introduced for efficiently handling multi-Deep Neural Network (DNN) tasks on heterogeneous embedded devices. This approach leverages the architectural heterogeneity of new embedded systems to address the challenges of workload management in modern edge data centers, which simultaneously handle multiple DNNs. RankMap employs stochastic space exploration combined with a performance estimator to navigate the extensive solution space of multi-DNN mapping. Experimental results demonstrate that RankMap achieves an average throughput that is 3.6 times higher than existing methods while preventing DNN starvation under heavy workloads and improving prioritization by 57.5 times. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary RankMap is a new way to manage many different neural networks on special devices called embedded systems. These devices are used in things like smart homes or self-driving cars. Right now, it’s hard for these devices to handle all the different networks at once. RankMap makes it easier by using random searches and good guesses about how well each network will do. This helps the device give priority to the most important networks and make everything run faster. |
Keywords
» Artificial intelligence » Neural network
