Résultats de recherche pour « HNSW »

Hierarchical Navigable Small World (HNSW) graphs serve as the industry standard due to their logarithmic complexity and strong empirical performance. However, HNSW relies on greedy graph traversal, a heuristic that provides no theoretical guarantees of correctness. In this paper, we propose a novel "Certify-then-Rectify" framework that bridges the gap between the speed of heuristic search and the rigor of exact retrieval. Rather than discarding HNSW, our approach first employs a distribution-free statistical certifier to dynamically evaluate the quality of a standard HNSW search with minimal overhead. If certification indicates that the retrieved neighbors are of low quality, the framework safely escalates to a rigorous exact recovery algorithm. To make this exact recovery computationally feasible, we reinterpret the HNSW graph as a geometric spanner and utilize Extreme Value Theory to stochastically estimate its maximum empirical stretch factor. This allows us to mathematically bound the maximum distance of true nearest neighbors. Extensive evaluations on benchmark datasets demonstrate that our tiered framework delivers the average-case speed of HNSW while ensuring the worst-case correctness of exact search and outperforming other applicable approaches.

The rapid growth of Large Language Models (LLMs) has generated massive amounts of high-dimensional feature vectors extracted from diverse datasets. Efficient storage and retrieval of such data are critical for enabling accurate and fast query responses. Vector databases (Vector DBs) provide efficient storage and retrieval for high-dimensional vectors. These systems rely on Approximate Nearest Neighbor Search (ANNS) indexes, such as HNSW, to handle large-scale data efficiently. However, the original HNSW is implemented on DRAM, which is both costly and vulnerable to crashes. Therefore, we propose P-HNSW, a crash-consistent HNSW on persistent memory. To guarantee crash consistency, P-HNSW introduces two logs, NLog and NlistLog. We describe the logging process during the operation and the recovery process in the event of system crashes. Our experimental results demonstrate that the overhead of the proposed logging mechanism is negligible, while P-HNSW achieves superior performance compared with SSD-based recovery mechanisms.

Vector databases (VectorDBs) are a core component of modern retrieval systems, including Retrieval-Augmented Generation (RAG), where efficient Approximate Nearest Neighbor Search (ANNS) is critical. Among ANNS algorithms, Hierarchical Navigable Small World (HNSW) graphs are widely adopted for their strong recalllatency trade-off. However, configuring HNSW remains challenging: its hyperparameters jointly affect search quality, latency, build time, and index size in nonlinear ways, while production deployments impose strict resource and tuning-time constraints.We study HNSW hyperparameter tuning from a systems perspective and show that its configuration space exhibits strong structural regularities. Specifically, we identify monotonic, dominant unimodal, and separable relationships among search-time and construction-time parameters, which induce feasibility boundaries under performance and resource constraints. Building on this insight, we propose CHAT, a constraint-aware tuning framework for HNSW. Unlike generic black-box optimizers, CHAT exploits HNSW-specific structure to perform deterministic, sample-efficient search and prune resource-infeasible configurations before full index construction. Across multiple datasets and HNSW-based vector search engines, CHAT identifies configurations that maximize recall or throughput while satisfying constraints on accuracy, latency, build time, index size, and tuning budget. Compared to strong baselines, CHAT achieves up to 45% higher throughput or 11% higher recall, and converges up to 44x faster. These results show that principled, structure-aware tuning enables efficient and robust HNSW deployment beyond generic black-box optimization.

Driven by recent breakthrough advances in neural representation learning, approximate near-neighbor (ANN) search over vector embeddings has emerged as a critical computational workload. With the introduction of the seminal Hierarchical Navigable Small World (HNSW) algorithm, graph-based indexes have established themselves as the overwhelmingly dominant paradigm for efficient and scalable ANN search. As the name suggests, HNSW searches a layered hierarchical graph to quickly identify neighborhoods of similar points to a given query vector. But is this hierarchy even necessary? A rigorous experimental analysis to answer this question would provide valuable insights into the nature of algorithm design for ANN search and motivate directions for future work in this increasingly crucial domain. We conduct an extensive benchmarking study covering more large-scale datasets than prior investigations of this question. We ultimately find that a flat navigable small world graph graph retains all of the benefits of HNSW on high-dimensional datasets, with latency and recall performance essentially \emph{identical} to the original algorithm but with less memory overhead. Furthermore, we go a step further and study \emph{why} the hierarchy of HNSW provides no benefit in high dimensions, hypothesizing that navigable small world graphs contain a well-connected, frequently traversed ``highway" of hub nodes that maintain the same purported function as the hierarchical layers. We present compelling empirical evidence that the \emph{Hub Highway Hypothesis} holds for real datasets and investigate the mechanisms by which the highway forms. The implications of this hypothesis may also provide future research directions in developing enhancements to graph-based ANN search.

Hierarchical Navigable Small World (HNSW) has demonstrated impressive accuracy and low latency for high dimensional nearest neighbor searches. However, its high computational demands and irregular, large-volume data access patterns present significant challenges to search efficiency. To address these challenges, we introduce P-HNSW, a software-hardware codesign solution for HNSW that integrates a Principal Component Analysis Filter (PCAF). On the software side, we apply PCAF to reduce dataset dimensionality, thereby lowering the volume of neighbor access and decreasing the computational load for distance calculations. On the hardware size, we design the P-HNSW processor with custom instructions to optimize search throughput and energy efficiency. Experimental results from RTL design synthesized using a 65nm technology node show that the proposed P-HNSW implementation achieves an 11.65× increase in Queries per Second (QPS) and a 56.8% reduction in energy consumption, compared to the original HNSW implementation

Efficient vector query processing is critical to enable AI applications at scale. Recent solutions struggle with growing vector datasets that exceed single-machine memory capacity, forcing unnecessary data movement and resource underutilization in monolithic architectures. We present d-HNSW, the first disaggregated vector similarity search engine for RDMA-based remote memory systems that achieves high performance while supporting fast data indexing with low network communication overhead. The core of d-HNSW is a novel disaggregation of the graph-based vector indexing data structure HNSW. It exploits the characteristics of greedy searching in HNSW to efficiently coordinate data transfers from the memory pool to the compute pool while serving data requests. Specifically, it leverages three ideas: (i) Representative index caching, a lightweight index constructed from a sampled subset of data, is cached in the compute pool to reduce frequent access to critical components of the hierarchical graph-based index, (ii) RDMA-friendly data layout design to reduce the networking round trips incurred by vector query and insertion and (iii) batched query-aware data loading to reduce bandwidth usage on data transfer between pools, addressing the limited cache capacity in compute nodes. We evaluate d-HNSW with extensive benchmarking datasets. The experimental results show that d-HNSW outperforms Naive d-HNSW implementation by up to 117x in latency while maintaining recall as 0.87 in dataset SIFT1M@1.

Approximate nearest neighbor (ANN) search is a fundamental problem in computer science for which in-memory graph-based methods, such as Hierarchical Navigable Small World (HNSW), perform exceptionally well. To scale beyond billions of high-dimensional vectors, the index must be distributed. The disaggregated memory architecture physically separates compute and memory into two distinct hardware units and has become popular in modern data centers. Both units are connected via RDMA networks that allow compute nodes to directly access remote memory and perform all the computations, posing unique challenges for disaggregated indexes. In this work, we propose a scalable HNSW index for ANN search in disaggregated memory. In contrast to existing distributed approaches, which partition the graph at the cost of accuracy, our method builds a graph-preserving index that reaches the same accuracy as a single-machine HNSW. Continuously fetching high-dimensional vector data from remote memory leads to severe network bandwidth limitations, which we overcome by employing an efficient caching mechanism. Since answering a single query involves processing numerous unique graph nodes, caching alone is not sufficient to achieve high scalability. We logically combine the caches of the compute nodes to increase the overall cache effectiveness and confirm the efficiency and scalability of our method in our evaluation.

The approximate nearest neighbor search (ANNS) is a fundamental and essential component in data mining and information retrieval, with graph-based methodologies demonstrating superior performance compared to alternative approaches. Extensive research efforts have been dedicated to improving search efficiency by developing various graph-based indices, such as HNSW (Hierarchical Navigable Small World). However, the performance of HNSW and most graph-based indices become unacceptable when faced with a large number of real-time deletions, insertions, and updates. Furthermore, during update operations, HNSW can result in some data points becoming unreachable, a situation we refer to as the `unreachable points phenomenon'. This phenomenon could significantly affect the search accuracy of the graph in certain situations. To address these issues, we present efficient measures to overcome the shortcomings of HNSW, specifically addressing poor performance over long periods of delete and update operations and resolving the issues caused by the unreachable points phenomenon. Our proposed MN-RU algorithm effectively improves update efficiency and suppresses the growth rate of unreachable points, ensuring better overall performance and maintaining the integrity of the graph. Our results demonstrate that our methods outperform existing approaches. Furthermore, since our methods are based on HNSW, they can be easily integrated with existing indices widely used in the industrial field, making them practical for future real-world applications. Code is available at \url{https://github.com/xwt1/MN-RU.git}