Tuesday, November 18, 2014

5 reasons to consider adopting IMC

big-data







This sponsored post is produced by GigaSpaces.


In-Memory Computing (IMC) is quickly becoming a must for processing and making sense of massive amounts of data. This is usually done by distributing the load across multiple servers and keeping things in memory, which results in massive performance improvements, and makes the difference between relevant, real-time insights and irrelevant pieces of information. In a recent webinar I did, a lot of questions came up about the common principles of IMC and why it should be implemented. So it made sense to give the answers to anyone considering IMC.




Now, it’s going to get a bit technical here, but I’ll try to keep it as high level as possible to explain why exactly. At its core, IMC involves performing common middleware functionalities, such as messaging, event processing, data access and transaction processing completely and exclusively in memory. If you’ve heard of such technologies as In-Memory Data Grids and In-Memory Databases, these are usually the enabling technologies behind IMC. By avoiding hard drives and minimizing network access, IMC delivers significantly faster performance.


The most important design element in IMC systems is focused on ensuring servers are as independent as possible from one another, often times referred to as “shared-nothing architecture.” This architecture is what makes these distributed systems extremely scalable, by distributing the load across multiple processes and servers.


This domain is no longer limited to exotic financial and telecom applications. Most, if not all, modern businesses render an ever-growing amount of services to their customers over the web, served using mobile devices, tablets and PCs. In such a business reality, application performance and scalability are mandatory, and real-time processing of data and events that flow into the business is a primary means to win over the competition.


5 reasons to consider adopting IMC


1. Memory is much faster than anything else out there: Random in-memory data access is up to 100 times faster than conventional hard drives.


2. Memory prices keep going down: In the past, memory was expensive and the amount of memory a 32-bit OS could manage was limited. With modern 64-bit operating systems, the amount of memory to which an application has access is no longer limited by software. On the hardware front, today you can buy half a terabyte of physical memory for the same price you would pay for less than 4GB of RAM 15 years ago. This means that a whole new set of problems can now be solved using only RAM.


3. When it comes to real time, you should measure throughput per dollar, not GB per dollar: Your business’s bottom line is dependent on performance and scalability of your applications. Slow processing results in immediate customer attrition; failure to serve the right content to the right customer — or worse, not serving content at all — will make you lose that customer. Inability to scale leads to downtime, which leads to loss of business and loss of reputation, which is far worse. And all that usually happens exactly when your business is growing.


4. Facebook, Netflix and most of Wall Street’s firms can’t be wrong: Wall Street firms are completely dependent on IMC to cope with competition, business growth and regulation. In-Memory Data Grids (IMDGs) are widely used across investment and retail banks, starting from order management systems in the front office and all the way to the back office to the reconciliation and settlement systems. Facebook uses in-memory in order to do real-time analytics, and Netflix just released their in-memory data distribution framework which is the foundation for the entire Netflix online experience.


5. It’s no longer bleeding edge, it’s actually pretty mainstream: IMC was one of Gartner’s top 10 technology trends for 2013, and more than a few large technology vendors have solid IMC offerings for over a decade. Products like GigaSpaces XAP (where I head the product team) have been serving mission-critical applications in production in the most demanding environments for a few years now, sitting at the heart of the business and realizing the promise of IMC.


Given all this, many are seeing IMC as a necessity now, rather than a luxury. Data is growing at an immeasurable speed, and the need to act upon that information is getting more important every day. IMC enables you to be able to process all of that data, in real-time — and really make it work for you.



Sponsored posts are content that has been produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. The content of news stories produced by our editorial team is never influenced by advertisers or sponsors in any way. For more information, contact sales@venturebeat.com.




Mobile developer or publisher? VentureBeat is studying mobile marketing automation.


Fill out our 5-minute survey, and we’ll share the data with you.



No comments: