Streaming JSON query engines achieve high throughputs by avoiding full document parsing and exploiting local parallelism. However, they struggle with point queries accessing only a small fraction of the data. Conversely, approaches based on a Document Object Model (DOM) excel at point queries through direct indexing, yet suffer from excessive memory usage and costly preprocessing. We propose a hybrid solution: a lightweight structural index based on Minimal Perfect Hash Functions that map opening bracket positions to their corresponding closing brackets, enabling efficient subtree skipping directly on the raw JSON stream. Thanks to our optimizations based on statistical analysis of real-world JSON datasets, the resulting index requires only a fraction of the JSON’s size. Our solution significantly accelerates query processing and closes the gap between streaming and DOM-based approaches. We achieve a mean speedup of 1.49 × and up to 20 × on point queries.

Publisher Copyright: © 2026 Copyright held by the owner/author(s)

Recently, graphs are becoming increasingly interesting in the context of large language models and as overlays for commercial databases. Subgraph query processing is an especially challenging workload for graph analysis that is bottlenecked by slow set intersection performance on CPUs. Previous work has shown the viability of utilizing hardware acceleration for related domains like graph and relational join processing. We propose GraphMatch, a hardware-accelerated subgraph query processing system based on worst-case optimal joins (WCOJ). For efficient processing of various data and query graphs, we propose a novel set intersection algorithm, called MaxStep, that leverages hardware parallelism. GraphMatch combines MaxStep operators in a data flow architecture which efficiently solves multi-set intersections in subgraph query processing, superior to CPU-based approaches. GraphMatch achieves an average speedup of over 6.98x and 17.08x, compared to the state-of-the-art WCOJ-based systems GraphFlow and RapidMatch, respectively. On labeled graphs, GraphMatch outperforms the fastest subgraph query processing accelerator FAST by orders of magnitude.

Research advancements in storage formats continuously produce more efficient encodings and better compression rates. Despite this, new formats are not adopted due to high implementation cost and existing formats cannot evolve because they need to maintain compatibility across systems. Can this problem be solved by introducing a new abstraction? We answer affirmatively with AnyBlox, a framework for reading arbitrary datasets using lightweight WebAssembly decoders bundled with the data. By decoupling decoders from both systems and file format specifications, AnyBlox allows transparent format evolution, instance-optimized encodings, and enables mainstream adoption of research advancements. It integrates seamlessly with modern systems like DuckDB, Spark, and Umbra, while delivering solid performance and security guarantees.

Data processing systems find themselves crushed between two moving tectonic plates: the usage plate driven by the system's users and their requirements; and the environment plate driven by various technological changes. We argue that the existing status quo of constantly adapting and thus bloating the system's implementation is simply unsustainable in the long run. We further argue that now is the right time to take a step back and establish the foundations of future-proof data processing systems that can easily adapt to different workloads and input formats, and that can run efficiently in any type of environment, today and in the future. With our paper, we analyze and learn from prior attempts, identify key design principles, and present our vision on how to design such systems.

The rising hardware heterogeneity in modern systems emphasizes new dimensions of optimizing task execution for data processing frameworks. Specialized hardware is often expected to be the exclusive executor of some particular workload because it was designed for it or is simply the fastest option. In heterogeneous database systems, almost always, the entire operation offloading is considered. However, little attention was given to database systems with horizontal cross-device pipeline parallelization. We argue that such an approach can be applied to systems with morsel-driven parallelism and improve performance. We apply our parallelization strategy to an existing system and accelerate aggregations using two devices by up to 1.5x compared to the fastest exclusive device executor.

This paper aims to bridge the gap between fast in-memory query engines and slow but robust engines that can utilize external storage. We find that current systems have to choose between fast in-memory operators and slower out-of-memory operators. We present a solution that leverages two independent but complementary techniques: First, we propose adaptive materialization, which can turn any hash-based in-memory operator into an out-of-memory operator without reducing in-memory performance. Second, we introduce self-regulating compression, which optimizes the throughput of spilling operators based on the current workload and available hardware. We evaluate these techniques using the prototype query engine Spilly, which matches the performance of state-of-the-art in-memory systems, but also efficiently executes large out-of-memory workloads by spilling to NVMe arrays.

Data science workloads frequently include Python code, but Python’s dynamic nature makes efficient execution hard. Traditional approaches either treat Python as a black box, missing out on optimization potential, or are limited to a narrow domain. However, a deep and efficient integration of user-defined Python code into data processing systems requires extracting the semantics of the entire Python code. In this paper, we propose a novel approach for extracting the high-level semantics by transforming general Python functions into program generators that generate a statically-typed IR when executed. The extracted IR then allows for high-level, domain-specific optimizations and the generation of efficient C++ code. With our prototype implementation, HiPy, we achieve single-threaded speedups of 2–20x for many workloads. Furthermore, HiPy is also capable of accelerating Python code in other domains like numerical data, where it can sometimes even outperform specialized compilers.

Chiplet-based CPUs, which combine multiple independent dies on a single package, allow hardware to scale to higher CPU core counts at the cost of more memory heterogeneity and performance variability. This introduces challenges when existing query engines are deployed on chiplet-based CPUs, as current designs make assumptions about uniform memory access, cache locality and consistent core performance, e.g., leading to ineffective CPU utilization. In this paper, we analyse the performance impact when query engines ignore chiplet-specific properties. We demonstrate that a naïve deployment can result in a significant degradation of query processing efficiency, exhibiting non-linear scaling even within a single CPU socket domain. Based on comprehensive experiments, we explore approaches to deploy query engines on chiplet-based CPUs with improved performance: we show that distributing processing tasks according to a chiplet-aware strategy achieves higher resource utilization and scalability, yielding an up to 7× speedup compared to hardware-oblivious approaches.

With full resource disaggregation on the horizon, it is unclear what the most suitable programming model is that enables dataflow developers to fully harvest the potential that recent hardware developments offer. In our vision, we propose to raise the abstraction level to allow developers to primarily reason about their dataflow and the requirements that need to be met by the underlying system in a declarative fashion. Underneath, the system works with typed memory regions and uses the notion of ownership that allows for more flexible memory management across the different compute devices and the tasks mapped onto them. This requires a holistic approach that crosses multiple layers of the system stack, opening exciting systems research questions.

Current hardware development greatly influences the design decisions of modern database systems. For many modern performance-focused database systems, query compilation emerged as an integral part and different approaches for code generation evolved, making use of standard compilers, general-purpose compiler libraries, or domain-specific code generators. However, development primarily focused on the dominating x86-64 server architecture; but neglected current hardware developments towards other CPU architectures like ARM and other RISC architectures. Therefore, we explore the design space of code generation in database systems considering a variety of state-of-the-art compilation approaches with a set of qualitative and quantitative metrics. Based on our findings, we have developed a new code generator called FireARM for AArch64-based systems in our database system, Umbra. We identify general as well as architecture-specific challenges for custom code generation in databases and provide potential solutions to abstract or handle them. Furthermore, we present an extensive evaluation of different compilation approaches in Umbra on a wide variety of x86-64 and ARM machines. In particular, we compare quantitative performance characteristics such as compilation latency and query throughput. Our results show that using standard languages and compiler infrastructures reduces the barrier to employing query compilation and allows for high performance on big data sets, while domain-specific code generators can achieve a significantly lower compilation overhead and allow for better targeting of new architectures.

Data processing systems face the challenge of supporting increasingly diverse workloads efficiently. At the same time, they are already bloated with internal complexity, and it is not clear how new hardware can be supported sustainably. In this paper, we aim to resolve these issues by proposing a unified abstraction layer based on declarative sub-operators in addition to relational operators. By exposing this layer to users, they can express their non-relational workloads declaratively with sub-operators. Furthermore, the proposed sub-operators decouple the semantic implementation of operators from the efficient imperative implementation, reducing the implementation complexity for relational operators. Finally, through fine-grained automatic optimizations, the declarative sub-operators allow for automatic morsel-driven parallelism. We demonstrate the benefits not only by providing a specific set of sub-operators but also implementing them in a compiling query engine. With thorough evaluation and analysis, we show that we can support a richer set of workloads while retaining the development complexity low and being competitive in performance even with specialized systems.

Despite the wide adoption of graph processing across many different application domains, there is no underlying data structure that can serve a variety of graph workloads (analytics, traversals, and pattern matching) on dynamic graphs with transactional updates. In this paper, we present Sortledton, a universal graph data structure that addresses the open problem by being carefully optimizing for the most relevant data access patterns used by graph computation kernels. It can support millions of transactional updates per second, while providing competitive performance (1.22x on average) for the most common graph workloads to the best-known baseline for static graphs - CSR. With this, we improve the ingestion throughput over state-of-the-art dynamic graph data structures, while supporting a wider range of graph computations under transactional guarantees, with a much simpler design and significantly smaller memory footprint (2.1x that of CSR).

Since its invention, data-centric code generation has been adopted for query compilation by various database systems in academia and industry. These database systems are fast but maximize performance at the expense of developer friendliness, flexibility, and extensibility. Recent advances in the field of compiler construction identified similar issues for domain-specific compilers and introduced a solution with MLIR, a generic infrastructure for domain-specific dialects. We propose a layered query compilation stack based on MLIR with open intermediate representations that can be combined at each layer. We further propose moving query optimization into the query compiler to benefit from the existing optimization infrastructure and make cross-domain optimization viable. With LingoDB, we demonstrate that the used approach significantly decreases the implementation effort and is highly flexible and extensible. At the same time, LingoDB achieves high performance and low compilation latencies.

Persistent memory (PMem) promised DRAM-like performance, byte addressability, and the persistency guarantees of conventional block storage. With the release of Intel Optane DCPMM, those expectations were dampened. While its write latency competes with DRAM, its read latency, write endurance, and especially bandwidth fall behind by up to an order of magnitude. Established PMem index structures mostly focus on lookups and cannot leverage PMem's low write latency. For inserts, DRAM-optimized index structures are still an order of magnitude faster than their PMem counterparts despite the similar write latency. We identify the combination of PMem's low write bandwidth and the existing solutions' high media write amplification as the culprit. We present Plush, a write-optimized, hybrid hash table for PMem with support for variable-length keys and values. It minimizes media write and read amplification while exploiting PMem's unique advantages, namely its low write latency and full bandwidth even for small reads and writes. On a 24-core server with 768 GB of Intel Optane DPCMM, Plush outperforms state-of-the-art PMem-optimized hash tables by up to 2.44X for inserts while only using a tiny amount of DRAM. It achieves this speedup by reducing write amplification by 80%. For lookups, its throughput is similar to that of established PMem-optimized tree-like index structures.

Moving data analysis and processing to the cloud is no longer reserved for a few companies with petabytes of data. Instead, the flexibility of on-demand resources is attracting an increasing number of customers with small to medium-sized workloads. These workloads do not occupy entire clusters but can run on single worker machines. However, picking the right worker for the job is challenging. Abstracting from worker machines, e.g., using stateless architectures, introduces overheads impacting performance. Solutions without stateless architectures resort to query restarts in the event of an adverse worker matching, wasting already achieved progress. In this paper, we propose migrating queries between workers by introducing on-demand state separation. Using state separation only when required enables maximum flexibility and performance while keeping already achieved progress. To derive the requirements for state separation, we first analyze the query state of medium-sized workloads on the example of TPC-DS SF100. Using this, we analyze the cost and describe the constraints necessary for state separation on such a workload. Furthermore, we describe the design and implementation of on-demand state separation in a compiling database system. Finally, using this implementation, we show the feasibility of our approach on TPC-DS and give a detailed analysis of the cost of query migration and state separation.

Dataflow graphs are a popular abstraction for describing computation, used in many systems for high-level optimization. For execution, dataflow graphs are lowered and optimized through layers of program representations down to machine instructions. Unfortunately, performance profiling such systems is cumbersome, as today's profilers present results merely at instruction and function granularity. This obfuscates the connection between profiles and high-level constructs, such as operators and pipelines, making interpretation of profiles an exercise in puzzling and deduction. In this paper, we show how to profile compiling dataflow systems at higher abstraction levels. Our approach tracks the code generation process and aggregates profiling data to any abstraction level. This bridges the semantic gap to match the engineer's current information need and even creates a comprehensible way to report timing information within profiling data. We have evaluated this approach within our compiling DBMS Umbra, showing that the approach is generally applicable for compiling dataflow systems and can be implemented with high accuracy and reasonable overhead.

An efficient implementation of a hash join has been a highly researched problem for decades. Recently, the radix join has been shown to have superior performance over the alternatives (e.g., the non-partitioned hash join), albeit on synthetic microbenchmarks. Therefore, it is unclear whether one can simply replace the hash join in an RDBMS or use the radix join as a performance booster for selected queries. If the latter, it is still unknown when one should rely on the radix join to improve performance.

A wealth of technology has evolved around relational databases over decades that has been successfully tried and tested in many settings and use cases. Yet, the majority of it remains overlooked in the pursuit of performance (e.g., NoSQL) or new functionality (e.g., graph data or machine learning). In this paper, we argue that a wide range of techniques readily available in databases are crucial to tackling the challenges the IT industry faces in terms of hardware trends management, growing workloads, and the overall complexity of a rapidly changing application and platform landscape. However, to be truly useful, these techniques must be freed from the legacy component of database engines: relational operators. Therefore, we argue that to make databases more flexible as platforms and to extend their functionality to new data types and operations requires exposing a lower level of abstraction: instead of working with SQL it would be desirable for database engines to compile, optimize, and run a collection of sub-operators for manipulating and managing data, offering them as an external interface. In this paper, we discuss the advantages of this, provide an initial list of such sub-operators, and show how they can be used in practice.

With today's data deluge, approximate filters are particularly attractive to avoid expensive operations like remote data/disk accesses. Among the many filter variants available, it is non-trivial to find the most suitable one and its optimal configuration for a specific use-case. We provide open-source implementations for the most relevant filters (Bloom, Cuckoo, Morton, and Xor filters) and compare them in four key dimensions: the false-positive rate, space consumption, build, and lookup throughput. We improve upon existing state-of-the-art implementations with a new optimization, radix partitioning, which boosts the build and lookup throughput for large filters by up to 9x and 5x. Our in-depth evaluation first studies the impact of all available optimizations separately before combining them to determine the optimal filter for specific use-cases. While register-blocked Bloom filters offer the highest throughput, the new Xor filters are best suited when optimizing for small filter sizes or low false-positive rates.

Multi-core scalability is one of the most important features for database systems running on today's hardware. Not surprisingly, the implementation of locks is paramount to achieving efficient and scalable synchronization. In this work, we identify the key database-specific requirements for lock implementations and evaluate them using both micro-benchmarks and full-fledged database workloads. The results indicate that optimistic locking has superior performance in most workloads due to its minimal overhead and latency. By complementing optimistic locking with a pessimistic shared mode lock we demonstrate that we can also process HTAP workloads efficiently. Finally, we show how lock contention can be handled gracefully without slowing down the uncontented fast path or increasing space requirements by using a lightweight parking lot infrastructure.

Implementing parallel operators in multi-core machines often involves a data partitioning step that divides the data into cache-size blocks and arranges them so to allow concurrent threads to process them in parallel. Data partitioning is expensive, in some cases up to 90% of the cost of, e.g., a parallel hash join. In this paper we explore the use of an FPGA to accelerate data partitioning. We do so in the context of new hybrid architectures where the FPGA is located as a co-processor residing on a socket and with coherent access to the same memory as the CPU residing on the other socket. Such an architecture reduces data transfer overheads between the CPU and the FPGA, enabling hybrid operator execution where the partitioning happens on the FPGA and the build and probe phases of a join happen on the CPU. Our experiments demonstrate that FPGA-based partitioning is significantly faster and more robust than CPU-based partitioning. The results open interesting options as FPGAs are gradually integrated tighter with the CPU.

In this paper we present BatchDB, an in-memory database engine designed for hybrid OLTP and OLAP workloads. BatchDB achieves good performance, provides a high level of data freshness, and minimizes load interaction between the transactional and analytical engines, thus enabling real time analysis over fresh data under tight SLAs for both OLTP and OLAP workloads.

For decades, database engines have found the generic interfaces offered by the operating systems at odds with the need for efficient utilization of hardware resources. As a result, most engines circumvent the OS and manage hardware directly. With the growing complexity and heterogeneity of modern hardware, database engines are now facing a steep increase in the complexity they must absorb to achieve good performance. Taking advantage of recent proposals in operating system design, such as multi-kernels, in this paper we explore the development of a light weight OS kernel tailored for data processing and discuss its benefits for simplifying the design and improving the performance of data management systems.

Efficient resource scheduling of multithreaded software on multicore hardware is difficult given the many parameters involved and the hardware heterogeneity of existing systems. In this paper we explore the efficient deployment of query plans over a multicore machine. We focus on shared query systems, and implement the proposed ideas using SharedDB. The goal of the paper is to explore how to deliver maximum performance and predictability, while minimizing resource utilization when deploying query plans on multicore machines. We propose to use resource activity vectors to characterize the behavior of individual database operators. We then present a novel deployment algorithm which uses these vectors together with dataflow information from the query plan to optimally assign relational operators to physical cores. Experiments demonstrate that this approach significantly reduces resource requirements while preserving performance and is robust across different server architectures.

Multicore computers pose a substantial challenge to infrastructure software such as operating systems or databases. Such software typically evolves slower than the underlying hardware, and with multicore it faces structural limitations that can be solved only with radical architectural changes. In this paper we argue that, as has been suggested for operating systems, databases could treat multicore architectures as a distributed system rather than trying to hide the parallel nature of the hardware. We first analyze the limitations of database engines when running on multicores using MySQL and PostgreSQL as examples. We then show how to deploy several replicated engines within a single multicore machine to achieve better scalability and stability than a single database engine operating on all cores. The resulting system offers a low overhead alternative to having to redesign the database engine while providing significant performance gains for an important class of workloads.