More Bang For Your Buck(et): Fast and Space-efficient Hardware-accelerated Coarse-granular Indexing on GPUs

2024-06-06

arXiv.org

AI Summary

Existing RTIndex (RX) suffers from three key bottlenecks when building database indexes on GPU ray-tracing hardware: high per-key memory overhead, slow range queries, and poor update efficiency. This paper proposes cgRX—the first RT Core-native indexing method designed specifically for coarse-grained bucket indexing. cgRX models keys as 3D triangular buckets and jointly optimizes space efficiency, query throughput, and dynamic update capability via serialized ray casting coupled with intra-bucket post-filtering. Compared to RX, cgRX achieves 1.5–3× higher memory-throughput efficiency, accelerates range queries by 2×, and performs single-key updates 5.5× faster than full index reconstruction. Crucially, cgRX is the first approach to simultaneously guarantee correctness and deliver performance gains for coarse-grained indexing on ray-tracing hardware.

Technology Category

• Search and Optimization: Distributed Search
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs
• Search and Retrieval-Augmented AI: Web crawling and indexing
• Security and Privacy: Large-scale security measurements

Abstract

In recent work, we have shown that NVIDIA's raytracing cores on RTX video cards can be exploited to realize hardware-accelerated lookups for GPU-resident database indexes. On a high level, the concept materializes all keys as triangles in a 3D scene and indexes them. Lookups are performed by firing rays into the scene and utilizing the index structure to detect hits in a hardware-accelerated fashion. While this approach called RTIndeX (or short RX) is indeed promising, it currently suffers from three limitations: (1) significant memory overhead per key, (2) slow range-lookups, and (3) poor updateability. In this work, we show that all three problems can be tackled by a single design change: Generalizing RX to become a coarse-granular index cgRX. Instead of indexing individual keys, cgRX indexes buckets of keys which are post-filtered after retrieval. This drastically reduces the memory overhead, leads to the generation of a smaller and more efficient index structure, and enables fast range-lookups as well as updates. We will see that representing the buckets in the 3D space such that the lookup of a key is performed both correctly and efficiently requires the careful orchestration of firing rays in a specific sequence. Our experimental evaluation shows that cgRX offers the most bang for the buck(et) by providing a throughput in relation to the memory footprint that is 1.5-3x higher than for the comparable range-lookup supporting baselines. At the same time, cgRX improves the range-lookup performance over RX by up to 2x and offers practical updateability that is up to 5.5x faster than rebuilding from scratch.

Problem

• Reduce memory overhead per key in GPU-resident database indexes
• Improve slow range-lookup performance in hardware-accelerated indexing
• Enhance updateability of the index structure for dynamic data

Innovation

• Hardware-accelerated indexing using GPU raytracing cores
• Coarse-granular indexing with key buckets for efficiency
• Optimized ray-firing sequence for correct and fast lookups