Open source 'critical' to big data for all
Technology vendors have backed a community-developed software platform as the critical piece to bring big data to the masses.
EMC and Intel have joined IBM in releasing their own flavours of an open-source software that manages how hardware in data centres accesses and processes information. The Hadoop open-source software by the Apache Software Foundation was released in October. It helps extract business insights from huge amounts of unstructured data, a trend commonly referred to as big data.
While it is free to download, its complexity has meant only large players, such as Google, Amazon, and Facebook, had the manpower, scale and incentive to customise it.
The new commercial releases are designed to make it easier, faster and safer to deploy. The vendors claim the cheaper technology will democratise big data.
This is Intel's first declaration of its software ambitions, according to Singapore-based Forrester analyst Dane Anderson, who said a standardised platform could unleash a new wave of data analytics innovation - like the iTunes App Store did for mobile apps.
"The risk for big data is that it becomes over-hyped when tech vendors just use it as a means to sell more of their gear," Anderson said.
"That isn't what big data is all about, it's about the services that allow you to extract business or social value. This is the building block for that to happen."
Famous for the chips inside personal computers, this time Intel has transcended the physical barrier to compete in the virtual world.
"The opportunity cost to not do this is that the evolution will be a lot slower, and the standards won't be established early enough," said Jason Fedder, Intel's general manager for Asia of data-centre products. "Intel's whole business model depends on open standards - it needs to make hundreds of millions of anything to make money - so it needs to accelerate the take up of those standards."
In 2004, EMC breached the software threshold when it acquired VMware, whose software generates additional server computing power, and became a preferred virtualisation platform.
The companies believe there is a similar opportunity in the big data market. For their troubles, they hope they will sell more of their Hadoop-friendly products.
Intel's announcement marks its first major foray into software. When asked whether it hoped to sell more Xeon processors from being the biggest contributor to the open-source project, Intel's mission critical computing director, Patrick Buddenbaum, said "absolutely".
The challenge for Intel, which failed to capitalise on the smartphone revolution, was to continue to drive demand for the hardware it made, he said.
Frequently Asked Questions about this Article…
Intel announced its first major move into software by backing a commercial distribution of the open‑source Hadoop platform. For investors, this matters because it signals a strategic shift from pure hardware to software-enabled services, which could expand Intel’s addressable market and influence demand for its data‑centre products.
Hadoop is an open‑source software platform from the Apache Software Foundation that helps organisations extract business insights from very large amounts of unstructured data. For investors, Hadoop is important because commercial distributions aim to make big data deployment easier and faster, potentially increasing adoption across many industries and creating opportunities for vendors that support the ecosystem.
The article notes Intel’s own team expects the company’s software involvement to drive more hardware sales—Intel’s mission‑critical computing director said the company hopes to sell more Xeon processors as a result. Supporting standardised, Hadoop‑friendly platforms could make it easier for customers to deploy big data workloads on Intel hardware, potentially lifting processor demand.
EMC and IBM (along with Intel) have released commercial Hadoop distributions to simplify the complexity of deploying open‑source Hadoop. The vendors say their releases make deployments easier, faster and safer, and they hope the resulting market growth will also increase sales of their Hadoop‑friendly products and services.
'Democratising big data' refers to making big data tools cheaper, easier and safer to use so more organisations—not just large tech players—can access analytics. For investors, broader adoption could translate into larger markets for software, hardware and services from vendors backing Hadoop distributions, but it’s not a guarantee of financial performance.
The article explains that while Hadoop is free to download, its complexity meant only large players like Google, Amazon and Facebook had the manpower, scale and incentive to customise and operate it effectively. Commercial distributions aim to lower that barrier for smaller organisations.
Analyst Dane Anderson warned that big data could become over‑hyped if technology vendors primarily use it to sell more hardware. He emphasised that big data should be about services that extract business or social value, not just a sales vehicle for equipment.
The article compares Intel’s software push to EMC’s 2004 acquisition of VMware, which moved EMC into software and virtualisation. This suggests a broader industry trend where hardware vendors diversify into software to establish standards, accelerate adoption and create new revenue streams—an important strategic consideration for investors assessing long‑term business models.
Call 1300 880 160
Start a chat
Schedule a call
