BCCE: Block-Centric GPU Co-Design for Real-Time Range-Top-K Query at Scale

Range-top-𝑘 queries retrieve the top-𝑘 elements within an arbitrary subrange of a large array and are a key primitive in real-time analytics. Unlike one-shot top-𝑘 selection, practical deployments issue large volumes of queries over varying and often overlapping ranges, frequently interleaved with streaming updates. In this setting, applying conventional GPU top-𝑘 kernels per query is inefficient: each query triggers range rescans or 𝑂 (𝑛)-scale passes that overwhelm This work is licensed under a Creative Commons Attribution 4.0 International License. HPDC '26, Cleveland, OH, USA