Flexible Deployment of Machine Learning Inference Pipelines in the Cloud–Edge–IoT Continuum

Electronics, Journal Year: 2024, Volume and Issue: 13(10), P. 1888 - 1888

Published: May 11, 2024

Currently, deploying machine learning workloads in the Cloud–Edge–IoT continuum is challenging due to wide variety of available hardware platforms, stringent performance requirements, and heterogeneity themselves. To alleviate this, a novel, flexible approach for inference introduced, which suitable deployment diverse environments—including edge devices. The proposed solution has modular design compatible with range user-defined pipelines. improve energy efficiency scalability, high-performance communication protocol propounded, along scale-out mechanism based on load balancer. service plugs into ASSIST-IoT reference architecture, thus taking advantage its other components. was evaluated two scenarios closely emulating real-life use cases, demanding requirements constituting several different scenarios. results from evaluation show that software meets high throughput low latency cases while effectively adapting hardware. code documentation, addition data used evaluation, were open-sourced foster adoption solution.

Language: Английский

---

Characterizing Dynamic Memory Behavior in WebAssembly Workloads

Published: May 5, 2024

Language: Английский

---

Wapplique: Testing WebAssembly Runtime via Execution Context-Aware Bytecode Mutation

Published: Sept. 11, 2024

Language: Английский

---

An Investigative Study of WebAssembly Performance in Cloud-to-Edge

Published: Sept. 21, 2024

Language: Английский

---