Welcome!

Java Authors: Elizabeth White, Yeshim Deniz, JP Morgenthal, Pat Romanski, Liz McMillan

Related Topics: Big Data Journal, SOA & WOA, Open Source, Virtualization, Cloud Expo, Apache

Big Data Journal: Article

Intel’s Going into the Hadoop Biz

Its reason for joining the Hadoop push is to sell more high-end Xeon server chips into scalable Hadoop clusters

Intel has gone into the open source Apache Hadoop business with its own distribution, which it calls the Intel Distribution for Apache Hadoop or simply the Intel Distribution.

Its reason for joining the Hadoop push is to sell more high-end Xeon server chips into scalable Hadoop clusters by goosing Hadoop's development. It also wants to move solid-state memory and its own networking.

Obviously it thinks the stuff is pretty fundamental.

As part of its effort Intel has added features to the Hadoop widgetry that nobody else has like silicon-based security. That makes Intel's the only Hadoop distribution to include complete encryption.

Since the widgetry supports the AES instructions in the Xeon chip, it's supposed to maximize Big Data performance and, together with the other improvements Intel has added, promises a 40% performance boost.

Intel's got 20 partners supporting the initial launch of its Hadoopery, as the Register calls it, a term we'll happily steal. These partners are supposed to integrate Intel's software into next-generation platforms and solutions and enable deployment in public and private cloud environments.

The partners include Cisco, Dell, Pentaho, Red Hat, SAP, SAS, Savvis, SuperMicro, Tableau, Teradata, Wipro and Zettaset.

Intel says it will open source all but it's most prized Hadoop platform enhancements as well as invest in more R&D to build analytic solutions for Hadoop. Whether this largesse gives rivals like Cloudera, MapR and Hortonworks a leg up remains to be seen. Intel's subscription pricing is supposed to be competitive.

Intel's off-limits proprietary software so far includes Intel Manager for Apache Hadoop software, which is supposed to simplify the deployment, configuration and monitoring of clusters for system administrators as they set up new applications.

Then there's Intel Active Tuner for Apache Hadoop software, which is supposed to take the guesswork out of performance tuning. Intel says until now this required a special understanding of each application's use of system resources along with Hadoop configuration and performance benchmarks. Now it doesn't.

Meanwhile, Intel's started Project Rhino, an open source effort to improve the data protection capabilities of the Hadoop ecosystem. The point is to improve encryption and authentication and make security more granular.

Intel imagines - and no one can gainsay this - that there will be literally tons of information coming from billions of sensors and intelligent systems; it estimates that the world generates a petabyte of data every 11 seconds, the equivalent of 13 years of HD video.

It thinks it can exploit the fact that "only a small fraction of the world is able to extract meaning from all of this information because the technologies, techniques and skills available today are either too rigid for the data types or too expensive to deploy."

And it figures the information derived from Hadoop can enrich our lives, not just make us easy marks for advertisers, by, say, accurately pinpointing customized treatments for terminal diseases.

Intel's distribution is supposed to analyze a terabyte of data, which normally takes more than four hours to fully process, in seven minutes because of its data-crunching chips and software.

Intel has been fooling around with its own Hadoop distribution for a few years ever since it formed a relationship with Yahoo and HP and is now on its third iteration based on the Apache Hadoop stack.

It has done work on the Hadoop Distributed File System, the YARN MapReduce 2.0 distributed processing framework, the Hive SQL query tool and the HBase key-value store.

Intel was reportedly pushed to putting something on the market next quarter by China Unicom and China Mobile, a couple of Chinese telecoms, to help with some performance issues in the Hadoop stack when running on Xeon chips.

Intel will distribute its version of Hadoop, developed in China, through vendors and service providers and sell its own technical support services. It says it won't fork the system.

More Stories By Maureen O'Gara

Maureen O'Gara the most read technology reporter for the past 20 years, is the Cloud Computing and Virtualization News Desk editor of SYS-CON Media. She is the publisher of famous "Billygrams" and the editor-in-chief of "Client/Server News" for more than a decade. One of the most respected technology reporters in the business, Maureen can be reached by email at maureen(at)sys-con.com or paperboy(at)g2news.com, and by phone at 516 759-7025. Twitter: @MaureenOGara

Comments (0)

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.