Flash for hedge fund

Leading European-based hedge fund adds Violin flash storage platform to support massive computational analyses on multi-terabyte datasets.

  • 8 years ago Posted in
A top European-based hedge fund has deployed the Violin 7300 Flash Storage Platform (FSP) for use in the computational analysis of massive datasets. The hedge fund selected Violin 7300 FSPs for their ability to support data persistence in analysing and back-testing trading strategies before running those strategies in live production.
This latest win by Violin involves adding 7300 FSPs to an existing Violin all-flash array deployment that supports the fund’s “calculation farm” which is used to increase the scope of the fund’s research zone. The research zone feeds data into the calculation farm and SQL databases — all of which have experienced significant performance improvements since the initial deployment of Violin all flash arrays.
For unified management of the arrays, the fund selected Violin’s Symphony flash management console, a single portal for managing petabytes of primary, flash-based storage across Violin’s flash storage platforms and all-flash arrays.
“Data persistence simply cannot be overlooked as a major contributor to success in the operations of high-performance, high-throughput applications that are critical to our operations,” said a spokesperson for the customer.  “Even when using multiple CPUs and parallel and grid computing techniques, we found that overall computational performance could be jeopardised as CPUs sat idle waiting for our previous disk-based arrays to catch up. The Violin all-flash array deployment had a sweeping effect on our operations, lowering latency and bringing overall performance increases that were astounding — ensuring improvements in data persistence. These performance improvements have influenced our decision to move forward to expand our flash storage environment to include Violin’s 7300 FSP.”
“Clients in the capital markets need to process existing and live data for a vast number of scenarios, which means they need an enterprise-class, persistent data platform that can concurrently read and write gigabytes of data per second — supporting the needs of multiple users to build and query datasets in realtime,” said Mike O’Hara, publisher of The Trading Mesh, an online magazine and social network for the global electronic trading industry.
“Saving milliseconds or even microseconds can provide quantum improvements in performing quantitative analysis on massive datasets,” said Kevin DeNuccio, president and chief executive officer, Violin Memory. “Quite simply, the place for legacy infrastructure, such as disk-based arrays is shrinking when flash storage platforms can deliver massive improvements in latency and in overall storage performance.”