Pivotal Launches With $105m Investment From GE
Startup company, Pivotal, announced today that they will be receiving a strategic investment from GE along the lines of $105 million dollars, representing a 10% equity stake in the company that was spun out of EMC to challenge the existing cloud software establishment.
Pivotal offers an enterprise platform as a service (PaaS) technology, dubbed Pivotal One, which the company bills as the first consumer-grade enterprise PaaS offering real-time, high capacity analytics in the cloud. The platform knits together three layers, a cloud fabric, a data fabric (which includes Pivotal’s own distribution of Hadoop which they call Pivotal HD), and an application fabric to provide businesses with the building blocks for creating big data applications and services.
“It’s no secret that the cloud and Big Data are driving dramatic business transformation enabling an industrial internet,” said Bill Ruh, Vice President and Corporate Officer of GE’s Global Software Center. “At the heart of it is that machines can be intelligent, connected, and that we can use software to analyze the information coming out of them.”
The investment fits in with GE’s vision of what they reference as “the Industrial Internet,” which they view as the next wave of connectivity encompassing intelligent machines, advanced analytics, and people at work, wherever they are. GE’s investment in Pivotal signals their intention to use Pivotal as the centerpiece for connecting and combining these elements.
“To support this next frontier requires an architectural shift in how our services are guilt and delivered,” commented Ruh. “Pivotal is creating a platform that brings the best of the Internet – rapid application development, data analytics and cloud architecture – to enterprises. This is aligned with many of the things we are doing at GE to help accelerate our delivery of innovation, and to bring a productivity revolution that will have a positive impact on all of us.”
Commenting on the market opportunity that exists Pivotal CEO, Paul Maritz explained that there is a need for a new class of real-time application emerging across very different industries that is not being serviced by the market today – a gap that Pivotal intends to fill.
Using the example of the modern Telco, Maritz explained that these companies largely exist in two different worlds that don’t communicate well with each other: the network world which delivers the calls over the air, and the customer support world.
“A Telco can tell you on average how many calls they’re dropping, but they can’t tell you in real time when the drop one of your calls,” commented Maritz in describing a relevant application for Pivotal. “Most importantly, when they drop one of your calls as a high value customer (or not), what are they going to do about that. To do that kind of thing, we have to bring together the ability to ingest large amounts of data in real time, hundreds of thousands of events per second, combine it with deep understanding of who we are dealing with, building analytic and statistical models over time, and then doing something in real time based upon that – not just sitting there looking and saying ‘ah we just dropped a call of a high value customer.’ You’ve got to do something about it in real time.
“So these kinds of applications we find across a wide variety of industries and when you see a set of applications emerging which there is no easy or cost effective way of doing that on existing infrastructure, that’s probably a strong hint that there is a new platform waiting to be born, and it’s that departure point that we at pivotal are going to try and be an example of a third generation platform.”
While the investment signals big promise for Pivotal’s future, they’ve still got a ways to go before they will be able to meet all the requirements of a company the size of GE, says Gartner Analyst Yefim Natis.
“Although the general direction of [Pivotal] is promising, some notable components that are essential to a comprehensive cloud computing platform are missing – there is no integration technology” commented Natis. “Integration of data, applications, cloud and Web services, partners and event streams is an essential element of any such environment. Pivotal will find that its customers demand that capability.”
Further, Natis comments that Pivotal currently lacks a high-productivity development platform, commenting that the foundation of Pivotal’s application platform, the CloudFoundry CEAP and Paas offer the Java or Ruby frameworks as the primary programming model. “A far cry in productivity from the cloud-native metadata-driven application PaaS (aPaas),” says Natis.
Noting that the first release of their Pivotal One platform is targeted for Q4 of this year, EMC and Pivotal Co-Founder, Scott Yara explained this morning during the Pivotal launch event that they’re not coming out with a “big bang” approach to solve every problem.
“The journey to executing the full vision of Pivotal One is really one that’s going to be measured in years, not months,” quipped Yara. “We are fairly pragmatic in understanding where we exist as a company and the kinds of things that we really need to do to live up to the hype.”
Part of that plan involves the implementation into Pivotal One of Pivotal’s own Hadoop distribution, which it calls Pivotal HD. Built on the core components of Apache Hadoop 2.0, Pivotal has added its own spin to the Hadoop framework that the company says it considers to be table stakes. Included in these unique flavorings is a proprietary Pivotal Command center that is designed to help users deploy, manage, and monitor the cluster and jobs within it. The distribution also includes a data loading mechanism that the company claims can load over 100 terabytes per hour. Pivotal also adds what it claims to be the first commercial Hadoop package to embed Hadoop virtualization extensions (HVE) from VMware, which Pivotal says allows their customers’ clusters to be physical node aware as well as virtual node aware.
Most notably, Pivotal’s unique Hadoop implementation integrates what they consider to be a key differentiator, HAWQ technology. Developed at EMC, they claim HAWQ brings more than 100x performance improvements to a wide range of query types and workloads (which is something that other open source efforts are currently wresting with). HAWQ is described as a massively parallel processing (MPP) database running on top of HDFS in Hadoop, embedded as one piece of converged infrastructure that extends Hadoop and HDFS in terms of performance and queryabilty. Notably, HAWQ is said to extend Hadoop by offering the full spectrum of the SQL interface, and by extension, the entire ecosystem of products that support SQL.
Of course, the proof of these claims will be in the pudding, but if true you can see where GE would take considerable interest in Pivotal to empower their industrial internet vision, which will be heavy in the use of telemetry where there is the potential of millions, even billions of devices phoning home to deliver vital data, among other use cases.