SOURCE: UBmatrix

May 07, 2008 12:45 ET

UBmatrix Sets the Standard for XBRL Processing Performance

Company Publicly Releases Performance Tests of XBRL Processing Engine for Use by XBRL Developers and IT Managers

REDWOOD CITY, CA and EINDHOVEN, THE NETHERLANDS--(Marketwire - May 7, 2008) - At the XBRL International Conference in Eindhoven, the Netherlands today, UBmatrix, Inc., the leading provider of XBRL information exchange solutions, released a set of performance data and standards documenting XBRL processing engine performance. As part of its public performance standards program, the company is offering for free a detailed whitepaper outlining standard and optimum XBRL processing performance including links to public download of XBRL test data and specifications.

"XBRL adoption is advancing at such a rapid pace and the XBRL development community has really been a model in terms of the sharing of information towards the common goal of advancing the XBRL standard. In that tradition, UBmatrix is publicly releasing XBRL processing engine performance data so that organizations and companies adopting XBRL can have a standard with which to measure performance," said Sunir Kapoor, Chief Executive Officer of UBmatrix, Inc. "There are so many variables that can impact performance, and until know developers have been operating largely in the dark in terms of having a measuring stick. We hope our public release of our data and learning will help in setting an industry standard."

The UBmatrix XBRL processing engine performance whitepaper summarizes the findings of an in-depth series of studies conducted by the company. The paper is intended for IT managers and developers who are developing XBRL platforms or adding XBRL capabilities to their applications environment. The whitepaper offers developers and IT managers a straightforward methodology for accurate performance testing of XBRL processing engines and detailed testing results. The UBmatrix XBRL Processing Engine Version 3.1 was used for the testing and the standards are based on tests with publically available sample XBRL taxonomies and instance documents. The testing included scenarios covering a range of real-world instance documents used commonly in the U.S. and Europe and were obtained from third-parties including XBRL-US and the Committee of European Banking Supervisors (CEBS). Links to all taxonomies and sample instance documents are included in the whitepaper for others wishing to run similar tests.

The whitepaper and presentation also include test methodologies to facilitate a more systematic approach to performance testing. The whitepaper outlines a methodology which includes:

--  Real-world simulations and a variety of performance testing scenarios;
--  Stress test scenarios to evaluate the impact of excessively large and
    complex documents;
--  Comparison of performance impact of single- and multi-threading CPUs;
--  Granular performance measurements to aid in engine tuning.

The methodology offers a way to produce an 'apples-to-apples' performance validation that considers the many factors that can impact processing speeds, as well as introduces opportunities for tuning engines to optimize performance in a given environment.

With the qualification that results vary based on the complexity of the instance document (report filing) and the number of facts reported, the study showed an average validation time of under 60 milliseconds per document. For further illustration, one test of a new US-GAAP taxonomy filing (552 facts reported and 7 contexts) showed scaling to over 10,000 documents per hour with other documents, such as a European COREP filing (235 facts reported and 9 context) averaging 25,000 documents per hour.

"When selecting an XBRL processing engine, it is vital that organizations analyze the various factors that impact processing time, understand how to optimize and test the engine in an environment to assess actual performance," said Steve Levine, Chief Marketing Officer, UBmatrix, Inc. "We're proud to provide the detailed performance data to help organizations configure their environments. UBmatrix's solutions are used in the most demanding XBRL environments and set the bar of XBRL processing performance."

For further information or to download the XBRL processing engine performance whitepaper, visit

UBmatrix Processing Engine

Now shipping, the UBmatrix Processing Engine 3.1, is the core platform for the development and deployment of XBRL-based applications. UBmatrix Processing Engine 3.1 is the market's highest performing processing engine. Version 3.1 raises the industry standards for XBRL platforms with the addition of powerful multithreading and streaming support. The UBmatrix Processing Engine is designed for high volume, large scale business processing and provides taxonomy and instance document validation and conversion which allow for transformation of XBRL from and into Excel, CVS, XML or HTML.

The UBmatrix Processing Engine is built on an extensible architecture and enables developers to easily integrate XBRL processing into Windows or Linux-based applications. The processing engine fully conforms to the latest XBRL standards and also supports formulas in anticipation of the upcoming XBRL International standard.

For further information on the business applications and value of XBRL processing engines, download a free whitepaper at

About UBmatrix

UBmatrix, Inc. is the leading provider of XBRL-based information exchange solutions for global organizations and enterprises allowing them to more efficiently and effectively address the challenges of business and financial information management, exchange and reporting. XBRL is fast becoming the standard for financial information exchange and reporting demonstrated by the XBRL programs being driven by the SEC, FDIC and the European Central Bank Supervisors. UBmatrix XBRL solutions increase operational efficiency and financial transparency, and ensure reporting accuracy and regulatory compliance. UBmatrix is based in Silicon Valley with development centers in Kirkland, WA, and New Delhi, India. For further information, visit

Contact Information

  • For further information, contact:

    Dan Cahill
    (917) 617-0106 Mobile
    Email Contact