skip to main content

GreenButton wants to become the omnipresent, defacto standard for speeding up computer hungry tasks, using the vast distributed computing power of the internet.

While the concept of this isn’t really new, commercialising has been the challenge. That is, the raw power is available as such, making it useable is where GB is making some inroads

Currently, infrastructure providers primarily sell resources based on time, and if you know exactly what you want, it's an easy game but what if you don’t?

For instance, render farms sell both the time and the license for processing, but bill afterwards.

It means users could end up with a massive bill and even during the actual processing people don’t know when the job will be completed. In these circumstances, with those risks, it is generally the desperate that engage.

The Wellington-based company has recently gone live with its patented 'Job Prediction Engine', which provides users at a desktop with a price estimate for different lengths of time a particular job can be carried out in.

Right Hemisphere’s Deep Exploration is the first cab of the rank to have a working version of the JP Engine.

With GB’ Job Prediction, a client decides what cost/speed option they want before they commit to the processing.

GB's job prediction algorithm combines user parameters, previous jobs and other factors to present price and time options to the client. Once decided by the user, the price is guaranteed and the tasks are then distributed across a multitude of global computers.

Since the processing is performed in the cloud, then other Internet workflow options are also available. This includes forwarding the completed job to YouTube for example, with no need to always have to download it back to the desktop.

Movie and design rendering, financial analysis, interpretation of complex oil and gas data or bioinformatic scrutiny that could take weeks on a single computer can arrive back within the time a client demands it.

They pay via credit card or PayPal. It is GB's micro-transaction ability that has attracted the attention of Microsoft and other large server farm vendors, who typically like to charge by the hour, but whose potential audience only wants to pony up for how much computer time they use.

GreenButton's team has been concentrating on its software development and its job prediction engine for the past eight months, attracting a consortium of investment from Movac, IceAngels, the NZ government and a number of its in-house members.

"This is new," says Vivian Morresey, GB's chief marketing officer.

In launching GB, in what is envisaged will be a business to business offering to software vendors who will integrate the application with other software packages, Morresey says "we have to go for a land grab."

"If we take too long, others who have more money and are quicker or faster, could take the market off us," he says. "However, since the software vendors will also share part of the fee charged for the cloud based processing, we are seeing more vendors run with us, instead of doing it themselves.”

In identifying and supplying a simple, cloud-based solution for people requiring ever more computing power, Morresey can envisage the time that users will not buy extra computers for processing - instead using GB to provide the IT muscle required, when they need it.

"We're part of the future of computing, which will be very cloud oriented," he says. "Our solution is simple and elegant, easy to use and intuitive. It means clients can carry out large computing tasks with certainty, relying on distributed computer power to deliver their requirements."

Morresey says that while GreenButton is making its play alongside software vendors, the company is open to other forms of ownership and collaboration with others.

In the meantime, as befits the constantly changing nature of the internet and relationships within it, GreenButton is working hard on becoming omnipresent he says.

Wellington start-up looks to become cloud computer crunching de facto standard

+ Text Size -
Original generation time 2.3539 seconds.