LIS25-313 Big Data & Data Science Project Update

• Friday, 16 May 11:10 - 11:35 (Europe/Madrid)
• Room: Session room 3 | Opala III

As computationally intensive workloads with high-density processing requirements, big data and data science applications have always been crucial deployment scenarios for ARM server architectures. These domains place exceptional demands on CPU computational capabilities, making them strategic priorities for ARM server ecosystem expansion. Through years of dedicated operation of the BDDS (Big Data and Data Science) project, Linaro has made substantial contributions to foundational open-source components including the data distribution Bigtop and the management tools Bigtop manager, we will elaborate on the progress and next stage plan for both two projects. Considering that vectorization as a critical path for CPU performance maximization for big data workloads, we have been actively exploring acceleration opportunities. We will share our progress in Spark SQL optimization through the Gluten vectorization framework and Velox execution engine. Besides, we will also share insights into potential future directions for big data infrastructure development

No slides available.