SCIENCE

Risk Management Solutions (RMS) has announced that it is to bring high-performance computing (HPC) to the industry through its new Enterprise Grid Computing software. Enterprise Grid Computing, which fully integrates the RMS RiskLink modeling platform with Microsoft's Windows HPC Server, allows computing resources to be used much more efficiently across the organization, with more flexible job and resource management and faster analysis run times providing higher productivity and deeper insight into modeled results.

Enterprise Grid Computing is an optional and separately licensed component of the RMS product suite to be released with version 10.0 of RMS' market-leading catastrophe modeling software. Major technology and infrastructure advancements in the latest version of RMS RiskLink and RiskBrowser will also deliver performance and scalability improvements, significantly increasing efficiency and reducing model runtimes.

Working closely with Microsoft, RMS is the first catastrophe modeling firm to have developed a solution specifically designed for a high-performance computing environment. By constructing large-scale computing grids that overcome bottlenecks for high volumes of data and analytics, Enterprise Grid Computing enables insurers and reinsurers to match their computing resources 'on demand' to users and jobs that create the biggest impact on the business. With Enterprise Grid Computing, run times can be significantly reduced, allowing companies to gain more up-to-date views of portfolio risk and a near real-time understanding of modeled losses. Advanced job management capabilities also allow the most critical jobs to preempt lower priority jobs that are already running.

"High-performance computing is revolutionizing the way companies do business by providing them with insights and analytics that are driving their decision-making," said Kyril Faenov, general manager of technical computing at Microsoft Corp. "The velocity of information is increasing, and the way that companies choose to capture and use it, by applying sophisticated analytics, will be the basis of future competition."

With Enterprise Grid Computing insurers and reinsurers will also be able to run more analytical cycles and process higher quantities of data than ever before, allowing them to gain unprecedented insights into their portfolios at the point of decision-making and test assumptions used in RMS models, as well as in their own portfolio exposure data.

"Catastrophe modeling is at a stage in its evolution where companies want to dig into sensitivity testing and gain a firm grasp of the models' assumptions, limitations, and capabilities," said Paul VanderMarck, chief products officer at RMS. "As advances in science and data increase the granularity and precision of models, high-performance computing accommodates the mounting volume of information and enables companies to gain insight into, and importantly, to quantify key drivers of uncertainty and make more informed decisions."

RMS recently won the 2010 Innovator Award by Windows in Financial Services magazine for its innovative use of Microsoft Windows HPC Server in its Enterprise Grid Computing software. The release of Enterprise Grid Computing is the initial element of a comprehensive suite of next-generation catastrophe modeling, portfolio management, and underwriting solutions from RMS.
 
Version 10.0 of the RMS Modeling Platform

The capabilities that Enterprise Grid Computing offers are in addition to the significant advancements in both RiskLink and RiskBrowser 10.0. With major technological and infrastructure improvements in this software, users will also benefit from a significant increase in performance, scalability and manageability of catastrophe risk analysis. Through increased performance in RMS' cat modeling engines and doubling the amount of computing power that can be effectively applied to any individual analysis, model runtimes can be reduced by 5 times in many cases, depending on the peril and region being analyzed. RiskLink 10.0 can also analyze portfolios of up to 5 million locations -- almost twice as many as version 9.0.

"RMS' new software solutions represent a sea-change in the way models will be used going forwards," commented Stephen Manning, managing director at Canopius. "The ability to harness advances in technology to maximize resource efficiencies and 'flex' the models is what the industry has been crying out for and is the next logical step for catastrophe modeling."