Actian Vector: A High-Performance Path Away From Netezza
This bank is one of the world’s largest financial institutions, providing individuals, small- to mid-sized businesses, and large public, private, and governmental institutions with a full range of banking, investing, asset, and risk management services. More than 50 million consumers and small businesses interact with the bank’s 5,000 retail banking offices and 16,000 ATMs. Online banking has about 30 million active users as well as more than 15 million mobile users.
When a key component of a global banking solution arrives at its end of life, a number of questions arise. How best to replace the key functionality? What more might a new solution enable the bank to deliver? How to migrate existing data sets to a new solution? What processes and requirements have changed over time and how best to invest now for future requirements?
All these questions whirled in the air when one global bank learned that neither IBM nor its channel partners planned to continue support for Netezza Twinfin, the analytics solution that this bank had relied on for years. Answers soon coalesced and the bank’s requirements became clear: they wanted a solution supporting a single data repository for all positions, across all asset classes. This would facilitate ad hoc analysis of positions and their sensitivity to market factors such as volatility, interest rate changes, timing, and more.
Additionally, the bank wanted a solution that would enable it to ingest, analyze, and act on massive volumes of both structured and unstructured data in short order. It had long been wanting deeper and more timely insights into its risk exposure (Netezza provided risk and opportunity value analyses only once per day via batch data dump), so the bank set its sights on a solution that could deliver insights in sub-minute intervals throughout the day.
And, of course, questions of performance, scalability, reliability, and cost were always present. The bank wanted the flexibility to build new analytical solutions as well as ease of management and maintenance to keep total cost of ownership (TCO) as low as possible. The solution needed to be secure yet not so rigid that it would preclude scaling with the bank’s needs over time.
After reviewing numerous alternatives and performing the requisite due diligence, the bank chose the Actian Vector Analytics Database and the Actian Data Platform™ to replace Netezza Twinfin. Actian Vector is a true column-store database that provides unrivaled abilities to connect, analyze, and act on big data. It turns even commodity hardware into a massively parallel high-performance data processing and real-time analytics solution capable of processing complex queries entirely in memory and returning actionable insights up to 30 times faster than competing solutions.
Indeed, as part of its search process the bank compared the performance of Actian Vector against Netezza Twinfin. Actian Vector, running on a compute cluster with only 5 nodes, performed on par with Netezza—which ran on 24-node compute cluster. The bank also performed a security audit and compared performance values relating to data compression, volume handling ability, load times, stress, and ad hoc query responses. Actian Vector outperformed its rivals in all these areas.
The bank also appreciated the flexibility and scalability ensured by Actian Vector. Actian Vector runs on commodity hardware on top of Windows or Linux. Adding nodes to a compute cluster is trivial in comparison to what the bank was used to with Netezza.
Sealing the deal was Actian’s level of expertise and commitment to partnership. Working with Actian, the bank saw a seamless pathway leading from Netezza to Actian Vector, which ultimately made the decision to embrace Actian Vector an easy one.
Today, the bank relies on Actian Vector to analyze both structured and unstructured data. It can run complex analytics in record time—gaining insights in moments that previously took hours or days. As anticipated from the proof of concept tests, the Actian solution has outperformed competing solutions in nearly every category, including the critically important time-to-value category.
Because Actian Vector can deliver extraordinary performance using only a small number of commodity compute nodes, the solution has exceeded the performance and functionality benchmarks of Netezza while lowering overall cost of ownership. By replacing its legacy technology, the bank estimates it will save $20 million over five years.
Has the bank encountered any downside to a decision that seemed so easy to make? Evidently not: the outcomes have been so satisfactory that the bank has already set its sights on incorporating Actian Vector into its consumer business and securities groups.