arXiv subjects

Bailu Ding

Publications and source records attributed to Bailu Ding.

9 recordsLinked to original sources

arXiv2026

ZipFlow: a Compiler-based Framework to Unleash Compressed Data Movement for Modern GPUs

In GPU-accelerated data analytics, the overhead of data transfer from CPU to GPU becomes a performance bottleneck when the data scales beyond GPU memory capacity due to the limited PCIe bandwidth. Data compression has come to rescue for reducing the amount of data transfer while taking advantage of the powerful GPU computation for decompression. To optimize the end-to-end query performance, however, the workflow of data compression, transfer, and decompression must be holistically designed based on the compression strategies and hardware characteristics to balance the I/O latency and computational overhead. In this work, we present ZipFlow, a compiler-based framework for optimizing compressed data transfer in GPU-accelerated data analytics. ZipFlow classifies compression algorithms into three distinct patterns based on their inherent parallelism. For each pattern, ZipFlow employs generalized scheduling strategies to effectively exploit the computational power of GPUs across diverse architectures. Building on these patterns, ZipFlow delivers flexible, high-performance, and holistic optimization, which substantially advances end-to-end data transfer capabilities. We evaluate the effectiveness of ZipFlow on industry-standard benchmark, TPC-H. Overall, ZipFlow achieves an average improvement of 2.08 times over the state-of-the-art GPU compression library (nvCOMP) and 3.14 times speedup against CPU-based query processing engines (e.g., DuckDB).

cs.DB↗

arXiv2025

RetroInfer: A Vector Storage Engine for Scalable Long-Context LLM Inference

Recent large language models (LLMs) are rapidly extending their context windows, yet inference throughput lags due to increasing GPU memory and bandwidth demands. This is because the key-value (KV) cache, an intermediate structure storing token representations, grows linearly with context length and requires an iterative linear scan for attention computation. A promising direction to accelerate long-context inference is to exploit attention's inherent sparsity by offloading the KV cache to CPU memory and retrieving only a small subset of tokens important to the current generation step. However, prior sparse attention approaches struggle to balance accuracy and retrieval cost due to varying sparsity patterns and inefficient GPU-CPU memory management. We present RetroInfer, a vector storage engine that realizes a sparsity-based KV cache for long-context inference. RetroInfer introduces an Attention-aWare VEctor index (wave index), which fundamentally improves the tradeoff between attention accuracy and retrieval cost through tripartite attention approximation, accuracy-bound attention estimation, and segmented clustering. We also design the wave buffer, a GPU-CPU buffer manager that assigns computation and manages data across heterogeneous hardware. We evaluate RetroInfer across a range of models and workloads, demonstrating up to 4.4X decoding throughput over full attention at 120K context and up to 12.2X over sparse attention baselines at 1 million tokens -- all while preserving full-attention-level accuracy.

cs.LG↗

arXiv2025

MINT: Multi-Vector Search Index Tuning

Vector search plays a crucial role in many real-world applications. In addition to single-vector search, multi-vector search becomes important for multi-modal and multi-feature scenarios today. In a multi-vector database, each row is an item, each column represents a feature of items, and each cell is a high-dimensional vector. In multi-vector databases, the choice of indexes can have a significant impact on performance. Although index tuning for relational databases has been extensively studied, index tuning for multi-vector search remains unclear and challenging. In this paper, we define multi-vector search index tuning and propose a framework to solve it. Specifically, given a multi-vector search workload, we develop algorithms to find indexes that minimize latency and meet storage and recall constraints. Compared to the baseline, our latency achieves 2.1X to 8.3X speedup.

cs.DB↗

arXiv2024

RetrievalAttention: Accelerating Long-Context LLM Inference via Vector Retrieval

Transformer-based Large Language Models (LLMs) have become increasingly important. However, due to the quadratic time complexity of attention computation, scaling LLMs to longer contexts incurs extremely slow inference speed and high GPU memory consumption for caching key-value (KV) vectors. This paper proposes RetrievalAttention, a training-free approach to both accelerate attention computation and reduce GPU memory consumption. By leveraging the dynamic sparsity of attention mechanism, RetrievalAttention proposes to build approximate nearest neighbor search (ANNS) indexes for KV vectors in CPU memory and retrieve the most relevant ones through vector search during generation. Unfortunately, we observe that the off-the-shelf ANNS indexes are often ineffective for such retrieval tasks due to the out-of-distribution (OOD) between query vectors and key vectors in the attention mechanism. RetrievalAttention addresses the OOD challenge by designing an attention-aware vector search algorithm that can adapt to the distribution of query vectors. Our evaluation demonstrates that RetrievalAttention achieves near full attention accuracy while only requiring access to 1--3% of the data. This leads to a significant reduction in the inference cost of long-context LLMs, with a much lower GPU memory footprint. In particular, RetrievalAttention only needs a single NVIDIA RTX4090 (24GB) to serve 128K tokens for LLMs with 8B parameters, which is capable of generating one token in 0.188 seconds.

cs.LG↗

arXiv2024

Retrieval with Learned Similarities

Retrieval plays a fundamental role in recommendation systems, search, and natural language processing (NLP) by efficiently finding relevant items from a large corpus given a query. Dot products have been widely used as the similarity function in such tasks, enabled by Maximum Inner Product Search (MIPS) algorithms for efficient retrieval. However, state-of-the-art retrieval algorithms have migrated to learned similarities. These advanced approaches encompass multiple query embeddings, complex neural networks, direct item ID decoding via beam search, and hybrid solutions. Unfortunately, we lack efficient solutions for retrieval in these state-of-the-art setups. Our work addresses this gap by investigating efficient retrieval techniques with expressive learned similarity functions. We establish Mixture-of-Logits (MoL) as a universal approximator of similarity functions, demonstrate that MoL's expressiveness can be realized empirically to achieve superior performance on diverse retrieval scenarios, and propose techniques to retrieve the approximate top-k results using MoL with tight error bounds. Through extensive experimentation, we show that MoL, enhanced by our proposed mutual information-based load balancing loss, sets new state-of-the-art results across heterogeneous scenarios, including sequential retrieval models in recommendation systems and finetuning language models for question answering; and our approximate top-$k$ algorithms outperform baselines by up to 66x in latency while achieving >.99 recall rate compared to exact algorithms.

cs.IR↗

arXiv2020

Bitvector-aware Query Optimization for Decision Support Queries (extended version)

Bitvector filtering is an important query processing technique that can significantly reduce the cost of execution, especially for complex decision support queries with multiple joins. Despite its wide application, however, its implication to query optimization is not well understood. In this work, we study how bitvector filters impact query optimization. We show that incorporating bitvector filters into query optimization straightforwardly can increase the plan space complexity by an exponential factor in the number of relations in the query. We analyze the plans with bitvector filters for star and snowflake queries in the plan space of right deep trees without cross products. Surprisingly, with some simplifying assumptions, we prove that, the plan of the minimal cost with bitvector filters can be found from a linear number of plans in the number of relations in the query. This greatly reduces the plan space complexity for such queries from exponential to linear. Motivated by our analysis, we propose an algorithm that accounts for the impact of bitvector filters in query optimization. Our algorithm optimizes the join order for an arbitrary decision support query by choosing from a linear number of candidate plans in the number of relations in the query. We implement our algorithm in Microsoft SQL Server as a transformation rule. Our evaluation on both industry standard benchmarks and customer workload shows that, compared with the original Microsoft SQL Server, our technique reduces the total CPU execution time by 22%-64% for the workloads, with up to two orders of magnitude reduction in CPU execution time for individual queries.

cs.DB↗

arXiv2015

The Homeostasis Protocol: Avoiding Transaction Coordination Through Program Analysis

Datastores today rely on distribution and replication to achieve improved performance and fault-tolerance. But correctness of many applications depends on strong consistency properties - something that can impose substantial overheads, since it requires coordinating the behavior of multiple nodes. This paper describes a new approach to achieving strong consistency in distributed systems while minimizing communication between nodes. The key insight is to allow the state of the system to be inconsistent during execution, as long as this inconsistency is bounded and does not affect transaction correctness. In contrast to previous work, our approach uses program analysis to extract semantic information about permissible levels of inconsistency and is fully automated. We then employ a novel homeostasis protocol to allow sites to operate independently, without communicating, as long as any inconsistency is governed by appropriate treaties between the nodes. We discuss mechanisms for optimizing treaties based on workload characteristics to minimize communication, as well as a prototype implementation and experiments that demonstrate the benefits of our approach on common transactional benchmarks.

cs.DB↗

arXiv2009

Contact graphs of disk packings as a model of spatial planar networks

Spatially constrained planar networks are frequently encountered in real-life systems. In this paper, based on a space-filling disk packing we propose a minimal model for spatial maximal planar networks, which is similar to but different from the model for Apollonian networks [J. S. Andrade, Jr. et al., Phys. Rev. Lett. {\bf 94}, 018702 (2005)]. We present an exhaustive analysis of various properties of our model, and obtain the analytic solutions for most of the features, including degree distribution, clustering coefficient, average path length, and degree correlations. The model recovers some striking generic characteristics observed in most real networks. To address the robustness of the relevant network properties, we compare the structural features between the investigated network and the Apollonian networks. We show that topological properties of the two networks are encoded in the way of disk packing. We argue that spatial constrains of nodes are relevant to the structure of the networks.

cond-mat.stat-mech↗

arXiv2009

Structural and spectral properties of a family of deterministic recursive trees: Rigorous solutions

As one of the most significant models, the uniform recursive tree (URT) has found many applications in a variety of fields. In this paper, we study rigorously the structural features and spectral properties of the adjacency matrix for a family of deterministic uniform recursive trees (DURTs) that are deterministic versions of URT. Firstly, from the perspective of complex networks, we investigate analytically the main structural characteristics of DURTs, and obtain the accurate solutions for these properties, which include degree distribution, average path length, distribution of node betweenness, and degree correlations. Then we determine the complete eigenvalues and their corresponding eigenvectors of the adjacency matrix for DURTs. Our research may shed light in better understanding of the features for URT. Also, the analytical methods used here is capable of extending to many other deterministic networks, making the precise computation of their properties (especially the full spectrum characteristics) possible.

cond-mat.stat-mech↗