Michael Bradley, VP Operations & Technology Services EMEA, at Violin Memory
There is somewhat of a rising demand for real-time data processing and, for a growing number of businesses today, this ability is paramount to their success. The driving forces behind this are twofold: first, the increasing expectation for information to be readily available at our fingertips, such as a bank balance on a mobile phone or social media channels, has rendered real-time data delivery essential to businesses of all sorts. Second, companies need to conduct instant forecasting and trend analysis based on high volumes of complex data, and in no sector is this need for speed of delivery, reliability and actionable data greater than in finance and banking.
Instant data delivery is often the Holy Grail for a financial organisation where even a miniscule delay in this process can have an operational and even financial impact; yet many are still using disk-based storage platforms and are struggling to achieve the necessary performance. The right technology here is one that will allow users to carry out thousands of transactions simultaneously and very rapidly, for example pulling data from a multiplicity of sources to turn it into meaningful and actionable data. This requires vast datasets to be read from storage systems, analysed and placed back onto the storage in near real-time. Traditional spinning disks are simply too slow and more and more business in the financial space are finding that no matter how much additional capacity they provide, the performance is simply not there. As a result data queries are not carried out at the desired speed, with a detrimental effect on user productivity or customer experience.
Flash, or memory-based storage, on the other hand, has become increasingly appealing to banks, insurers and other financial organisations thanks to its blend of speed, reliability and Total Cost of Ownership (TCO). This technology is able to deliver 150 microseconds at 4GB/s and has a 95 per cent lower latency than disk drives, resulting in an accelerated analytics cycle, quicker insights and the ability to transform data from a technical problem into a business opportunity.
While analysing information has been mission critical to financial institutions for several decades, the volumes of data put under the microscope today are growing on an unprecedented scale. It is generally accepted that in a single day we create 2.5 quintillion bytes, or one trillion eight-hundred-and-forty billion GBs of data; so much that 90 per cent of the data in the world today was created in the last two years alone. Furthermore, businesses are not just dealing with structured datasets as they were in the past; instead, available information has become unstructured and far more complex and is often derived from numerous sources, rendering its interpretation evermore challenging.
Turning multiple and complex datasets, or Big Data, into meaningful information has become one of the biggest business challenges of our times. Companies within the financial markets for example, are continually seeking out solutions to assist them in understanding trading lifecycles, strategy execution and post-trade processing. These insights are critical to performance and staying ahead of the competition.
Analysing Big Data is essential for many organisations, but in the financial sector there is another mission-critical dimension: timing. Rapid interpretation of data, which is ever-growing in volume, the speed it is generated at and its types, requires new levels of performance unattainable with legacy technologies. Flash-based storage allows users to sort, analyse and store data in near real-time, meaning that even the largest volumes can be turned into meaningful information in a matter of hours, while they are still valuable, and well ahead of competitors relying on traditional disk drive technology. It’s astonishing to see how many legacy infrastructures built to cope with the demands of the last century are still in use today. Changing to new technologies can seem time consuming, resource intensive and expensive, but it can still make more sense than throwing more of the same (product) at a problem. Our requirements have changed and grown to meet the challenges presented by today’s business dynamics and so too has technology. As they say, there are no old paths to new destinations.
Leading Identity Authentication Solution Provider Replaces Traditional Disk with Violin Memory
Finsphere, a leading provider of identity authentication services to global financial institutions, required an innovative solution to accelerate ID authentication response rates to more accurately detect, prioritise and manage security challenges.
Finsphere’s identity authentication services address a critical challenge as digital and mobile transactions grow. Real-time identity authentication depends on fast response time from the underlying IT infrastructure and high application throughput. Financial institutions look to reduce the risk of payment card fraud by relying on ultra-low latency and network backbone access that meet strict performance requirements.
Violin Memory, which is outperforming disk-based storage arrays with a 28x higher throughput, 2.7x quicker response times, and almost 22x more IOPs, while also lowering cost, provided Finsphere with flash Memory Arrays. This enabled Finsphere’s real-time fraud alerts and identity authentication and helping the company to markedly improve response rates to financial institution customers and further differentiate itself in the marketplace.
Finsphere also found that the use of the Violin Flash Memory Array for its purpose-built application drastically reduced the cost per IOP by nearly nine times (from £8.88 to £1.03). Co-location data centre costs are also reduced significantly, as the Violin solution has a much smaller footprint and lower power requirements than traditional disk-based SANs. Benchmarks against the company’s existing provider already show nearly 22x more IOPs (38,339.22 v.1771.05), higher throughput (621.52MBps v. 22.70 MBps) and ultimately quicker response times (1.67ms v. 4.52ms).
Finsphere’s card protection solution leverages both financial data and telecom (mobile geo-location) data to verify the identity of the card user. Inconsistencies such as a credit card transaction in Paris while the legitimate cardholder’s mobile phone is located in London can trigger a real-time alert or other action to reduce the risk of fraud.
“I first read about Violin in a tech journal and the performance claims seemed too good to be true. So we benchmarked our application with Violin’s Flash Memory Arrays, and found they delivered on their promises,” said Mary Reeder, chief technology officer of Finsphere. “I was very impressed and genuinely excited at what this would mean for us. The performance is outstanding, and more than meets the demands of high-volume transaction processing environments like ours.”