Scanning HTML at Tens of Gigabytes per Second on ARM Processors [r-libre/3616]

Lemire, Daniel (sous presse). Scanning HTML at Tens of Gigabytes per Second on ARM Processors. Software: Practice and Experience.

Fichier(s) associé(s) à ce document :

PDF - simdhtml-6.pdf
Contenu du fichier : Document de travail
Licence : Creative Commons CC BY.

Catégorie de document :	Articles de revues
Évaluation par un comité de lecture :	Oui
Étape de publication :	Accepté (sous presse)
Résumé :	Modern processors have instructions to process 16 bytes or more at once. These instructions are called SIMD, for single instruction, multiple data. Recent advances have leveraged SIMD instructions to accelerate parsing of common Internet formats such as JSON and base64. During HTML parsing, they quickly identify specific characters with a strategy called vectorized classification. We review their techniques and compare them with a faster alternative. We measure a 20-fold performance improvement in HTML scanning compared to traditional methods on recent ARM processors. Our findings highlight the potential of SIMD-based algorithms for optimizing Web browser performance.
Déposant:	Lemire, Daniel
Responsable :	Daniel Lemire
Dépôt :	03 mars 2025 18:36
Dernière modification :	11 mars 2025 20:48

RÉVISER