HP Technology at WorkThe must-read IT business eNewsletter
Meet your need for speed
More, faster, better—these are the demands most enterprises are facing today. They’re handling more data than ever before, and are expected to provide enough processing power to analyze it fast. This need for speed has brought new technologies like in-memory computing (IMC) to the forefront, as enterprises explore better ways to handle their compute demands.
IMC can provide such benefits as super-fast data caching, storing, and processing. Unlike traditional databases that rely on disk storage mechanisms, IMC uses main memory for storing data. Based on simplified algorithms, IMC executes fewer CPU instructions and reduces seek-time during the data-query process, resulting in much faster performance than disk-based data storage.
The impact of IMC: data consolidation and processing speed
By reducing processes that once took hours or minutes down to minutes or seconds, IMC can offer real-time and near-real-time services that previously weren’t possible.
“There are two aspects that in-memory computing can address,” says Frances Guida, manager of Converged Systems Messaging and Content at HP. “Businesses have a lot of data that gets distributed in many different databases across many different disks—and they have a hard time making use of all that information the way it’s distributed around the organization.” In-memory computing allows enterprises to make better use of that information, she says, as users “can analyze massive data sets much more quickly and with much more ease. You can put the power of all that data together and handle it in a real-time fashion—for example, performing business analytics on your production data without impacting your ability to process transactions.”
The second part of the equation is the speed that IMC imparts on analysis and Big Data sets. “By having everything in-memory, by definition, you’re always getting the speed that allows you to do things like take transaction times that might have taken minutes, and move them to seconds,” Guida says. Queries and analytical workloads that could take hours can be reduced “to seconds, because you don’t have to go out to that slow disk and can access the data straight out of the memory.”
HP ConvergedSystem for SAP HANA
One of the things customers are challenged by when they look at a technology like in-memory computing is how much time and effort to invest. Companies “know how to design an architecture to support traditional databases, but may not know much about in-memory computing,” Guida says. “Customers are discovering that their core competence lies less in the technical aspects of engineering servers, storage, and networking and more in how they can deliver faster MRP reports, for example.”
That’s where HP ConvergedSystem for SAP HANA comes in. HP has “hundreds of engineers who work on SAP on a daily basis,” Guida explains, “and we have intimate knowledge of how to design servers, storage, and networking.” Customers get the performance, availability, and scalability they need from IMC without having to build and engineer it themselves. As a common platform designed for faster Big Data analytics, ConvergedSystem for SAP HANA can provide real-time, actionable business insights and efficiencies.
“We have companies that have taken those classic MRP processes that used to take hours and get that down to minutes or seconds,” Guida says. “For example, India’s largest bicycle manufacturer used to take one and a half to two hours to run their MRP reports. They had to do them overnight because the system didn’t enable them to run during the day. Now it takes only 10 to 15 minutes, and they can run reports several times per day.”
HP Solutions for SAP HANA
HP Opens New Center of Excellence for In-memory Computing