Improvements include access to eXtremeDB Financial Edition's most powerful capabilities from the widely-used SQL and Python languages, unleashing productivity by simplifying development and, as a result, reducing costs and slashing time-to-deployment. Version 6.0 also tackles the challenge of Big Data in financial systems with new features including distributed query processing and specialized market data compression, to streamline handling of real-time and historical databases that today grow beyond one trillion records.

"In STAC-M3 benchmarks and other tests, eXtremeDB Financial Edition has shown it can deliver the lowest latency in processing market data. Version 6.0 builds on this core strength, with some improvements accelerating performance, others increasing cost-effectiveness, and several new features doing both. Major companies in capital markets are already leveraging the improvements in eXtremeDB Financial Edition 6.0. McObject looks forward to helping additional customers tackle their toughest financial computing challenges as the upgrade goes into general release," said Steve Graves, McObject co-founder and CEO.

McObject will be presenting performance results using eXtremeDB Financial Edition 6.0 at 2014 STAC Summit conferences in London (October 30), Chicago (November 4) and New York (November 13).

Graves provided details on the improvements in eXtremeDB Financial Edition 6.0 as follows:

Vector-based statistical functions in SQL, Python.

A defining feature of eXtremeDB Financial Edition is its library of vector-based statistical functions, which are pipelined in CPU cache to minimize latency when analyzing market data. In version 6.0, the product introduces the ability to use its vector-based functions from within SQL, a database language that is nearly universal in enterprise computing. This vastly increases the pool of developers whose existing skills permit quick adoption of McObject's database system.

SQL can be used in applications written in languages including C/C++, Python, Java and C# (.NET) and also on its own, via a command-line interface. Its high level of productivity for creating code that interacts with database systems enables faster development in a field where timing can be a key competitive advantage. McObject's video gives a practical example of using vector-based functions from within SQL.

eXtremeDB Financial Edition 6.0 also adds support for Python, a high-level language favored in capital markets because of its fast time-to-deployment, particularly for tasks such as rapid prototyping. Using Python alongside eXtremeDB Financial Edition's dynamic database definition language (DDL) capability, developers can implement their ideas quickly, and optimize rapidly by testing changes to code, database tables and indexes.

Distributed query processing.

With this new capability, eXtremeDB Financial Edition partitions a database and distributes query processing across multiple servers, CPUs and/or CPU cores. Performance is accelerated — dramatically, in some cases — via parallel execution of database operations and by harnessing the capabilities of many host computers rather than just one.

Market data compression.

The upgrade adds a powerful new compression algorithm for reducing the size of stored market data, to reduce storage costs and accelerate processing. This run-length encoding (RLE) compression can be applied to columnar data (i.e. fields defined as eXtremeDB Financial Edition's 'sequence' data type used for trades, quotes and other market data). In McObject's tests, activating this feature reduced Chicago Board Options Exchange Market Volatility Index (VIX) tick data to one-quarter of its pre-compression size, and increased speed in reading the database by 21%. (eXtremeDB Financial Edition 6.0 also adds a feature for compressing non-columnar data.)

Improvements to vector-based statistical function library.

eXtremeDB Financial Edition 6.0 improves its vector-based function library in additional ways, including:
• Many new functions, including a family of hash aggregate functions that lend efficiency in processing sequences that are unordered (i.e. not sorted by timestamp or fixed interval)
• The ability to create a sequence from an input string, as in the following example, which creates the sequence named "bid": "insert into Quote (bid) values ('{1.0, 1.1, 1.2, 1.1}')"
• The ability, via a new SQL SELECT statement keyword, to "flatten" the results of queries on columnar data (i.e. time series data such as market data) into a row-based format that is more familiar to users of traditional relational database systems