article
For some time now, web browsers have sought to achieve native desktop performance when hosting and executing Compute-Intensive Applications (CIAs) within their Execution Environment (EE). Various technologies have come and gone to that end, but none of them have really sustainably bridged the performance gap that exists between web browser EEs and the native desktop. One obstacle to web browser's hosting and executing CIAs with native desktop performance is the lack of web browser technologies with the necessary architecture and design to support them. This paper seeks to briefly recap the current state of web browser EEs, specifically where they relate to performance as based on our previously published work. We then propose a novel solution dubbed System23 (SYS23), that has been architectured and designed to bring the performance of applications executing in a web browser EE much closer to the performance of a native desktop. We then present a preliminary set of findings based on performance benchmarks that have been collected using the PolyBench/C benchmarking suite. Preliminary results indicate that the architecture and design of SYS23 holds up quite well against existing EEs from a performance point of view. Where the SYS23 performance is within 1% of that of a native desktop and at least 30% faster than today's fastest web browser EE, namely the WebAssembly Programming Language (WASM). Based on our preliminary work, we will look to complete the full study after which we aim to publish our full set of findings to the community for further analysis and subsequent further investigation. Limitations that have already surfaced surround the fact that this solution only focuses on the web browser EE and does not concern itself with the building or compiling of the web browser applications nor with the distribution mechanisms of these applications. Those are some areas that can be highlighted as requiring further investigation in future studies.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1109/icodse63307.2024.10829873
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.