RNTuple performance: Status and Outlook

Preprints and early-stage research may not have been peer reviewed yet.
To read the file of this research, you can request a copy directly from the authors.


Upcoming HEP experiments, e.g. at the HL-LHC, are expected to increase the volume of generated data by at least one order of magnitude. In order to retain the ability to analyze the influx of data, full exploitation of modern storage hardware and systems, such as low-latency high-bandwidth NVMe devices and distributed object stores, becomes critical. To this end, the ROOT RNTuple I/O subsystem has been designed to address performance bottlenecks and shortcomings of ROOT's current state of the art TTree I/O subsystem. RNTuple provides a backwards-incompatible redesign of the TTree binary format and access API that evolves the ROOT event data I/O for the challenges of the upcoming decades. It focuses on a compact data format, on performance engineering for modern storage hardware, for instance through making parallel and asynchronous I/O calls by default, and on robust interfaces that are easy to use correctly. In this contribution, we evaluate the RNTuple performance for typical HEP analysis tasks. We compare the throughput delivered by RNTuple to popular I/O libraries outside HEP, such as HDF5 and Apache Parquet. We demonstrate the advantages of RNTuple for HEP analysis workflows and provide an outlook on the road to its use in production.

No file available

Request Full-text Paper PDF

To read the file of this research,
you can request a copy directly from the authors.

ResearchGate has not been able to resolve any citations for this publication.
Full-text available
The ROOT TTree data format encodes hundreds of petabytes of High Energy and Nuclear Physics events. Its columnar layout drives rapid analyses, as only those parts (“branches”) that are really used in a given analysis need to be read from storage. Its unique feature is the seamless C++ integration, which allows users to directly store their event classes without explicitly defining data schemas. In this contribution, we present the status and plans of the future ROOT 7 event I/O. Along with the ROOT 7 interface modernization, we aim for robust, where possible compile-time safe C++ interfaces to read and write event data. On the performance side, we show first benchmarks using ROOT’s new experimental I/O subsystem that combines the best of TTrees with recent advances in columnar data formats. A core ingredient is a strong separation of the high-level logical data layout (C++ classes) from the low-level physical data layout (storage backed nested vectors of simple types). We show how the new, optimized physical data layout speeds up serialization and deserialization and facilitates parallel, vectorized and bulk operations. This lets ROOT I/O run optimally on the upcoming ultra-fast NVRAM storage devices, as well as file-less storage systems such as object stores.
Full-text available
A new event data format has been designed and prototyped by the CMS collaboration to satisfy the needs of a large fraction of physics analyses (at least 50%) with a per event size of order 1 kB. This new format is more than a factor of 20 smaller than the MINIAOD format and contains only top level information typically used in the last steps of the analysis. The talk will review the current analysis strategy from the point of view of event format in CMS (both skims and formats such as RECO, AOD, MINIAOD, NANOAOD) and will describe the design guidelines for the new NANOAOD format.
Full-text available
The analysis of High Energy Physics (HEP) data sets often takes place outside the realm of experiment frameworks and central computing workflows, using carefully selected "n-tuples" or Analysis Object Data (AOD) as a data source. Such n-tuples or AODs may comprise data from tens of millions of events and grow to hundred gigabytes or a few terabytes in size. They are typically small enough to be processed by an institute's cluster or even by a single workstation. N-tuples and AODs are often stored in the ROOT file format, in an array of serialized C++ objects in columnar storage layout. In recent years, several new data formats emerged from the data analytics industry. We provide a quantitative comparison of ROOT and other popular data formats, such as Apache Parquet, Apache Avro, Google Protobuf, and HDF5. We compare speed, read patterns, and usage aspects for the use case of a typical LHC end-user n-tuple analysis. The performance characteristics of the relatively simple n-tuple data layout also provides a basis for understanding performance of more complex and nested data layouts. From the benchmarks, we derive performance tuning suggestions both for the use of the data formats and for the ROOT (de-)serialization code.
Conference Paper
High level abstractions in Python that can utilize computing hardware well seem to be an attractive option for writing data reduction and analysis tasks. In this paper, we explore the features available in Python which are useful and efficient for end user analysis in High Energy Physics (HEP). A typical vertical slice of an HEP data analysis is somewhat fragmented: the state of the reduction/analysis process must be saved at certain stages to allow for selective reprocessing of only parts of a generally time-consuming workflow. Also, algorithms tend to to be modular because of the heterogeneous nature of most detectors and the need to analyze different parts of the detector separately before combining the information. This fragmentation causes difficulties for interactive data analysis, and as data sets increase in size and complexity (O10 TiB for a "small" neutrino experiment to the O10 PiB currently held by the CMS experiment at the LHC), data analysis methods traditional to the field must evolve to make optimum use of emerging HPC technologies and platforms. Mainstream big data tools, while suggesting a direction in terms of what can be done if an entire data set can be available across a system and analysed with high-level programming abstractions, are not designed with either scientific computing generally, or modern HPC platform features in particular, such as data caching levels, in mind. Our example HPC use case is a search for a new elementary particle which might explain the phenomenon known as "Dark Matter". Using data from the CMS detector, we will use HDF5 as our input data format, and MPI with Python to implement our use case.
ROOT RNTuple Virtual Probe Station
  • J Blomer
J. Blomer et al. ROOT RNTuple Virtual Probe Station. Accessed: 2022-02-10. url:
ROOT-An object oriented data analysis framework". In: Nuclear instruments and methods in physics research section A: accelerators, spectrometers, detectors and associated equipment 389
  • Rene Brun
  • Fons Rademakers
Rene Brun and Fons Rademakers. "ROOT-An object oriented data analysis framework". In: Nuclear instruments and methods in physics research section A: accelerators, spectrometers, detectors and associated equipment 389.1-2 (1997), pp. 81-86.
vmtouch: the Virtual Memory Toucher
  • Hoyte Doug
Hoyte Doug. vmtouch: the Virtual Memory Toucher. 2012. url: