Parallelism and WebAssembly memory management in microinterface architecture
Information Technology and Computer Engineering · 03.08.2026 · (23) · 121-131
DOI: 10.31649/vitce/2.2026.121 ↗Kurz gefasst
Fachartikel in der Zeitschrift Information Technology and Computer Engineering, online erschienen am 03.08.2026. Verfasst von Oleksandr Stepanov und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The decentralisation of monolithic front-ends into micro-front-end architectures, combined with WebAssembly (Wasm), imposes limitations on the communication overhead associated with ‘bridge crossing’ between JavaScript and Wasm’s linear memory, an area that remains under-researched in quantitative terms. The aim of the study was to quantitatively assess the overhead of bridge crossing, analyse the risks of race conditions in shared-memory micro-frontend environments, and develop and validate a hybrid communication model that eliminates serialisation bottlenecks. A controlled micro-benchmark methodology was employed for the study, utilising the browser API performance.now() with a measurement precision of no less than 5 microseconds. Each scenario was run over 10,000 measurement iterations following a mandatory JIT compiler warm-up phase of 1,000 iterations. Two architecturally distinct data transfer strategies were compared under identical hardware conditions using Float32Array arrays ranging in size from 0.1 MB to 10 MB at a rendering rate of 60 frames per second. It has been empirically established that the transfer of primitive data types takes 50-100 nanoseconds, whereas copying large data arrays (e.g., 1 MB) takes 1-3 milliseconds – which, at a frame rate of 60 frames per second, consumes 11.1% of available CPU time solely on data transfer, completely negating Wasm’s performance advantages. The proposed hybrid model, which combines a lightweight event bus for transmitting control signals with a SharedArrayBuffer and Atomics-based synchronisation for data access, achieves a transaction time of 250 nanoseconds regardless of payload size – corresponding to a speed-up factor of up to 66,000 times for 10 MB arrays compared to classical event-driven communication. The results provided a quantitatively validated architectural foundation for designing high-performance loosely coupled micro-frontend systems capable of performing AI/ML inference, 3D rendering and video streaming directly within the browser without architectural compromises
Zitieren
APA
Stepanov, O., & Klym, H. (2026). Parallelism and WebAssembly memory management in microinterface architecture. Information Technology and Computer Engineering, 121-131. https://doi.org/10.31649/vitce/2.2026.121
Vancouver
Stepanov O, Klym H. Parallelism and WebAssembly memory management in microinterface architecture. Information Technology and Computer Engineering. 2026:121-131. doi:10.31649/vitce/2.2026.121
BibTeX
@article{Stepanov2026,
author = {Stepanov, Oleksandr and Klym, Halyna},
title = {Parallelism and WebAssembly memory management in microinterface architecture},
journal = {Information Technology and Computer Engineering},
year = {2026},
number = {23},
pages = {121--131},
doi = {10.31649/vitce/2.2026.121}
}
EUVIMED