arXiv subjects

Carsten Binnig

Publications and source records attributed to Carsten Binnig.

At least 19 recordsLinked to original sources

arXiv2026

Correctness Firewalls for Synthesized Database Components

For the first time in history, high-performance code is cheap. Today's coding agents can generate bespoke, instance-optimized data systems with excellent performance, and such agents will continue to improve as LLMs advance. But a purely agentic approach, even when validated with strong testing techniques, does not automatically inherit the correctness guarantees our community has spent decades building. We propose a middle ground between hand-crafted and agent-crafted systems: correctness firewalls, lightweight design patterns under which a synthesized component can affect a system's performance but never its results. Behind a correctness firewall, agents are free to synthesize fast components while the overall system remains correct. We present three case studies, each demonstrating a different type of firewall: cardinality estimation (advisory), searching sorted data (guarded), and data compression (certified). In early experiments, synthesized components search sorted data up to 3x faster than human-designed learned indexes and can compress text 1.5x smaller than Zstd. Drawing on these case studies, we outline a research agenda for how data management systems can exploit coding agents without giving up their hard-won correctness guarantees, from protecting synthesized code to measuring the cost of correctness.

cs.DB↗

arXiv2026

The Stretto Execution Engine for Semantic Data Systems

Semantic data systems enable querying over both structured and unstructured data using LLMs, but a central challenge is navigating the performance-accuracy trade-off when selecting among alternative strategies for executing semantic operators. In this paper, we present Stretto, a new execution engine for semantic query processing that combines a novel optimizer with a rich set of physical execution strategies for semantic operators. Unlike prior work, Stretto provides global - rather than per-operator - accuracy guarantees while optimizing query performance. To this end, Stretto formulates query planning as a constrained optimization problem and employs a gradient-based optimizer to jointly select operator implementations and allocate error budgets across the full query plan. Furthermore, to enable diverse execution choices, Stretto introduces a new technique that builds on KV-cache compression to realize a spectrum of physical execution strategies for semantic operators with controllable runtime-accuracy trade-offs. Experimental results demonstrate that Stretto outperforms state-of-the-art systems while consistently satisfying quality guarantees.

cs.DB↗

arXiv2026

From Data Querying to Data Investigations: Rethinking Natural Language Interfaces for Databases

Natural language (NL) interfaces to databases have been optimized for the wrong problem. The dominant Text-to-SQL paradigm assumes that users ask questions that can be answered by single SQL queries. In practice, however, users seek assistance with solving data problems. This requires searching a database by sequences of SQL queries while reasoning over intermediate results instead of just running one SQL query. This paper therefore introduces a new paradigm for NL interfaces to data, which we call data investigations. We present D^2, a first prototype of a data investigation system that embodies this vision by autonomously searching, reasoning over, and collecting data to solve data problems. Using a newly constructed benchmark based on the Murder Mystery dataset, we demonstrate the potential of D^2 for tasks that require data investigations with evidence-backed decisions, extending beyond the capabilities of traditional single-query question answering.

cs.DB↗

arXiv2026

TEmBed-T: A Multi-Dimensional Benchmark for Table-Level Embeddings

Tabular data is the dominant structured-data modality, and learning table representations has become a core research direction. Table-level embeddings in particular underpin a wide range of applications, including table retrieval, data lake discovery, and table classification. Despite their importance, there is still limited understanding of how different embedding approaches behave across tasks, making systematic evaluation and analysis essential. In this work, we introduce a systematic evaluation of table-level embeddings that captures several complementary properties required for downstream effectiveness. We realize this evaluation by extending TEmBed, a recently proposed testbed for tabular embeddings, whose table-level coverage is currently limited to a single retrieval task. An empirical study over the TEmBed model pool confirms that no single model excels across all tasks, demonstrating that table-level embedding quality cannot be reduced to retrieval alone.

cs.DB↗

arXiv2026

High-Performance DBMSs with io_uring: When and How to use it

We study how modern database systems can leverage the Linux io_uring interface for efficient, low-overhead I/O. io_uring is an asynchronous system call batching interface that unifies storage and network operations, addressing limitations of existing Linux I/O interfaces. However, naively replacing traditional I/O interfaces with io_uring does not necessarily yield performance benefits. To demonstrate when io_uring delivers the greatest benefits and how to use it effectively in modern database systems, we evaluate it in two use cases: Integrating io_uring into a storage-bound buffer manager and using it for high-throughput data shuffling in network-bound analytical workloads. We further analyze how advanced io_uring features, such as registered buffers and passthrough I/O, affect end-to-end performance. Our study shows when low-level optimizations translate into tangible system-wide gains and how architectural choices influence these benefits. Building on these insights, we derive practical guidelines for designing I/O-intensive systems using io_uring and validate their effectiveness in a case study of PostgreSQL's recent io_uring integration, where applying our guidelines yields a performance improvement of 14%.

cs.DB↗

arXiv2026

Bespoke OLAP: Synthesizing Workload-Specific One-size-fits-one Database Engines

Modern OLAP engines are designed to support arbitrary analytical workloads, but this generality incurs structural overhead, including runtime schema interpretation, indirection layers, and abstraction boundaries, even in highly optimized systems. An engine specialized to a fixed workload can eliminate these costs and exploit workload-specific data structures and execution algorithms for substantially higher performance. Historically, constructing such bespoke engines has been economically impractical due to the high manual engineering effort. Recent advances in LLM-based code synthesis challenge this tradeoff by enabling automated system generation. However, naively prompting an LLM to produce a database engine does not yield a correct or efficient design, as effective synthesis requires systematic performance feedback, structured refinement, and careful management of deep architectural interdependencies. We present Bespoke OLAP, a fully autonomous synthesis pipeline for constructing high-performance database engines tightly tailored to a given workload. Our approach integrates iterative performance evaluation and automated validation to guide synthesis from storage to query execution. We demonstrate that Bespoke OLAP can generate a workload-specific engine from scratch within minutes to hours, achieving order-of-magnitude speedups over modern general-purpose systems such as DuckDB.

cs.DB↗

arXiv2026

Bespoke-Card: Why Tune When You Can Generate? Synthesizing Workload-Specific Cardinality Estimators

Cardinality estimators are built to support arbitrary schemas and workloads, forcing them to rely on generic statistics even when the schema and workload is known in advance, leaving optimizers prone to large errors and poor plans. We present Bespoke-Card, an agent-driven system that synthesizes workload-specific cardinality estimators as executable code: a planning agent designs the estimators' strategies, a coding agent implements them, and a validator scores the estimates against true cardinalities and PostgreSQL estimates, forming a robust and deterministic harness. Going beyond naive prompting, Bespoke-Card uses structured q-error feedback, regression analysis, concrete outlier subplans, a curriculum isolating join-only, filter-only, and full-subplan errors, and archival selection of the best implementation. Injecting its estimates into the optimizer cuts total PostgreSQL runtime on JOB by 33% and reduces median q-error over all JOB subplans from 190.5 to 11.5 (-94%), while synthesizing a strong estimator in under one hour for less than $10. Bespoke-Card is opening a new avenue for cardinality estimation next to classical generic estimators and learned estimator architectures.

cs.DB↗

arXiv2026

Selectivity Estimation for Semantic Filters on Image Data

Semantic data systems integrate Large Language Models (LLMs) and Vision-Language Models (VLMs) directly into database query execution, enabling expressive queries on multi-modal data. However, optimizing these queries requires accurate selectivity estimates to determine the most efficient operator execution order. Contemporary systems rely on online sample-based profiling, a process that incurs severe latency overheads and struggles with low-selectivity queries. In this paper, we introduce Semantic Histograms, a novel selectivity estimator for semantic filters on image data that leverages shared embedding spaces to bypass traditional profiling. We realize that all semantic filters are implicit range queries, as they match a range of different images. Some filter predicates are more general, yielding a wide range, while others are more specific, yielding a smaller range. To address the challenge of implicit ranges, we propose two approaches to estimate the queries' specificity, with an ensemble of the two performing best. The evaluation shows that Semantic Histograms can reduce the end-to-end runtime overhead of query optimization and execution by up to 86%.

cs.DB↗

arXiv2026

Redbench: Workload Synthesis From Cloud Traces

Workload traces from cloud data warehouse providers reveal that standard benchmarks such as TPC-H and TPC-DS fail to capture key characteristics of real-world workloads, including query repetition and string-heavy queries. In this paper, we introduce Redbench, a novel benchmark featuring a workload generator that reproduces real-world workload characteristics derived from traces released by cloud providers. Redbench integrates multiple workload generation techniques to tailor workloads to specific objectives, transforming existing benchmarks into realistic query streams that preserve intrinsic workload characteristics. By focusing on inherent workload signals rather than execution-specific metrics, Redbench bridges the gap between synthetic and real workloads. Our evaluation shows that (1) Redbench produces more realistic and reproducible workloads for cloud data warehouse benchmarking, and (2) Redbench reveals the impact of system optimizations across four commercial data warehouse platforms. We believe that Redbench provides a crucial foundation for advancing research on optimization techniques for modern cloud data warehouses.

cs.DB↗

arXiv2026

Do GPUs Really Need New Tabular File Formats?

Parquet is the de facto columnar file format in modern analytical systems, yet its configuration guidelines have largely been shaped by CPU-centric execution models. As GPU-accelerated data processing becomes increasingly prevalent, Parquet files generated with CPU-oriented defaults can severely underutilize GPU parallelism, turning GPU scans into a performance bottleneck. In this work, we systematically study how Parquet configurations affect GPU scan performance. We show that Parquet's poor GPU performance is not inherent to the format itself but rather a consequence of suboptimal configuration choices. By applying GPU-aware configurations, we increase effective read bandwidth up to 125 GB/s without modifying the Parquet specification.

cs.DB↗

arXiv2026

PystachIO: Efficient Distributed GPU Query Processing with PyTorch over Fast Networks & Fast Storage

The AI hardware boom has led modern data centers to adopt HPC-style architectures centered on distributed, GPU-centric computation. Large GPU clusters interconnected by fast RDMA networks and backed by high-bandwidth NVMe storage enable scalable computation and rapid access to storage-resident data. Tensor computation runtimes (TCRs), such as PyTorch, originally designed for AI workloads, have recently been shown to accelerate analytical workloads. However, prior work has primarily considered settings where the data fits in aggregated GPU memory. In this paper, we systematically study how TCRs can support scalable, distributed query processing for large-scale, storage-resident OLAP workloads. Although TCRs provide abstractions for network and storage I/O, naive use often underutilizes GPU and I/O bandwidth due to insufficient overlap between computation and data movement. As a core contribution, we present PystachIO, a prototype of a PyTorch-based distributed OLAP engine that combines fast network and storage I/O with key optimizations to maximize GPU, network, and storage utilization. Our evaluation shows up to 3x end-to-end speedups over existing distributed GPU-based query processing approaches.

cs.DB↗

arXiv2025

Unveiling Challenges for LLMs in Enterprise Data Engineering

Large Language Models (LLMs) promise to automate data engineering on tabular data, offering enterprises a valuable opportunity to cut the high costs of manual data handling. But the enterprise domain comes with unique challenges that existing LLM-based approaches for data engineering often overlook, such as large table sizes, more complex tasks, and the need for internal knowledge. To bridge these gaps, we identify key enterprise-specific challenges related to data, tasks, and background knowledge and extensively evaluate how they affect data engineering with LLMs. Our analysis reveals that LLMs face substantial limitations in real-world enterprise scenarios, with accuracy declining sharply. Our findings contribute to a systematic understanding of LLMs for enterprise data engineering to support their adoption in industry.

cs.DB↗

arXiv2025

Towards a Multimodal Stream Processing System

In this paper, we present a vision for a new generation of multimodal streaming systems that embed MLLMs as first-class operators, enabling real-time query processing across multiple modalities. Achieving this is non-trivial: while recent work has integrated MLLMs into databases for multimodal queries, streaming systems require fundamentally different approaches due to their strict latency and throughput requirements. Our approach proposes novel optimizations at all levels, including logical, physical, and semantic query transformations that reduce model load to improve throughput while preserving accuracy. We demonstrate this with Samsara, a prototype leveraging such optimizations to improve performance by more than an order of magnitude. Moreover, we discuss a research roadmap that outlines open research challenges for building a scalable and efficient multimodal stream processing systems.

cs.DB↗

arXiv2025

Reflex: Faster Secure Collaborative Analytics via Controlled Intermediate Result Size Disclosure

Secure Multi-Party Computation (MPC) enables collaborative analytics without exposing private data. However, OLAP queries under MPC remain prohibitively slow due to oblivious execution and padding of intermediate results with filler tuples. We present Reflex, the first framework that enables configurable trimming of intermediate results across different query operators -- joins, selections, and aggregations -- within full query plans. At its core is the Resizer operator, which can be inserted between any oblivious operators to selectively remove filler tuples under MPC using user-defined probabilistic strategies. To make privacy trade-offs interpretable, we introduce a new metric that quantifies the number of observations an attacker would need to infer the true intermediate result sizes. Reflex thus makes the performance-privacy space of secure analytics navigable, allowing users to balance efficiency and protection. Experiments show substantial runtime reductions while maintaining quantifiable privacy guarantees.

cs.DB↗

arXiv2025

Is it Bigger than a Breadbox: Efficient Cardinality Estimation for Real World Workloads

DB engines produce efficient query execution plans by relying on cost models. Practical implementations estimate cardinality of queries using heuristics, with magic numbers tuned to improve average performance on benchmarks. Empirically, estimation error significantly grows with query complexity. Alternatively, learning-based estimators offer improved accuracy, but add operational complexity preventing their adoption in-practice. Recognizing that query workloads contain highly repetitive subquery patterns, we learn many simple regressors online, each localized to a pattern. The regressor corresponding to a pattern can be randomly-accessed using hash of graph structure of the subquery. Our method has negligible overhead and competes with SoTA learning-based approaches on error metrics. Further, amending PostgreSQL with our method achieves notable accuracy and runtime improvements over traditional methods and drastically reduces operational costs compared to other learned cardinality estimators, thereby offering the most practical and efficient solution on the Pareto frontier. Concretely, simulating JOB-lite workload on IMDb speeds-up execution by 7.5 minutes (>30%) while incurring only 37 seconds overhead for online learning.

cs.DB↗

arXiv2025

Opening The Black-Box: Explaining Learned Cost Models For Databases

Learned Cost Models (LCMs) have shown superior results over traditional database cost models as they can significantly improve the accuracy of cost predictions. However, LCMs still fail for some query plans, as prediction errors can be large in the tail. Unfortunately, recent LCMs are based on complex deep neural models, and thus, there is no easy way to understand where this accuracy drop is rooted, which critically prevents systematic troubleshooting. In this demo paper, we present the very first approach for opening the black box by bringing AI explainability approaches to LCMs. As a core contribution, we developed new explanation techniques that extend existing methods that are available for the general explainability of AI models and adapt them significantly to be usable for LCMs. In our demo, we provide an interactive tool to showcase how explainability for LCMs works. We believe this is a first step for making LCMs debuggable and thus paving the road for new approaches for systematically fixing problems in LCMs.

cs.DB↗

arXiv2025

JOB-Complex: A Challenging Benchmark for Traditional & Learned Query Optimization

Query optimization is a fundamental task in database systems that is crucial to providing high performance. To evaluate learned and traditional optimizer's performance, several benchmarks, such as the widely used JOB benchmark, are used. However, in this paper, we argue that existing benchmarks are inherently limited, as they do not reflect many real-world properties of query optimization, thus overstating the performance of both traditional and learned optimizers. In fact, simple but realistic properties, such as joins over string columns or complex filter predicates, can drastically reduce the performance of existing query optimizers. Thus, we introduce JOB-Complex, a new benchmark designed to challenge traditional and learned query optimizers by reflecting real-world complexity. Overall, JOB-Complex contains 30 SQL queries and comes together with a plan-selection benchmark containing nearly 6000 execution plans, making it a valuable resource to evaluate the performance of query optimizers and cost models in real-world scenarios. In our evaluation, we show that traditional and learned cost models struggle to achieve high performance on JOB-Complex, providing a runtime of up to 11x slower compared to the optimal plans.

cs.DB↗

arXiv2025

PDSP-Bench: A Benchmarking System for Parallel and Distributed Stream Processing

The paper introduces PDSP-Bench, a novel benchmarking system designed for a systematic understanding of performance of parallel stream processing in a distributed environment. Such an understanding is essential for determining how Stream Processing Systems (SPS) use operator parallelism and the available resources to process massive workloads of modern applications. Existing benchmarking systems focus on analyzing SPS using queries with sequential operator pipelines within a homogeneous centralized environment. Quite differently, PDSP-Bench emphasizes the aspects of parallel stream processing in a distributed heterogeneous environment and simultaneously allows the integration of machine learning models for SPS workloads. In our results, we benchmark a well-known SPS, Apache Flink, using parallel query structures derived from real-world applications and synthetic queries to show the capabilities of PDSP-Bench towards parallel stream processing. Moreover, we compare different learned cost models using generated SPS workloads on PDSP-Bench by showcasing their evaluations on model and training efficiency. We present key observations from our experiments using PDSP-Bench that highlight interesting trends given different query workloads, such as non-linearity and paradoxical effects of parallelism on the performance.

cs.DC↗