Small start-ups have taken a lead on enterprises in their adoption of big data, by embedding analytics into the heart of their applications, according to Amazon Web Services.As the cloud has grown, prices for rentable computing power have come down, while scale-out applications like Hadoop have evolved to analyse large quantities of data, Werner Vogels, Amazon Web Service's (AWS) chief technology officer, said in a keynote speech at the AWS Summit 2012 in London.
Small start-ups have taken a lead on enterprises in their adoption of big data, by embedding analytics into the heart of their applications, according to Amazon Web Services.
As the cloud has grown, prices for rentable computing power have come down, while scale-out applications like Hadoop have evolved to analyse large quantities of data, Werner Vogels, Amazon Web Service's (AWS) chief technology officer, said in a keynote speech at the AWS Summit 2012 in London.
"What we would consider a level playing field is actually tilting towards younger businesses — they are using big data tools much more rapidly to improve the core of their businesses," Vogels said on Tuesday. "We hope the cloud will give [businesses] a level playing field so you can execute these programs just as easily."
Vogels pointed to the constellation of big data services within AWS, such as Amazon Elastic MapReduce, the just-released AWS Marketplace and the high-powered cluster compute rentable computers, as examples of how the cloud takes on the capital and operations costs associated with processing big data.
"The infrastructure piece of doing big data is substantial," he said. "Cloud has made big data easy."
By way of an example, he picked out Cycle Computing, which was able to create a 50,000-core supercomputing cluster in the AWS cloud for drug discovery specialist Schroedinger.
This supercomputer ran for around three hours at a cost of $4,900 (£3,033) per hour and analysed 21 million chemical compounds and whether they worked against a protein target. An equivalent physical supercomputer would cost $20m to build, Vogels said.
Cloud has made big data easy.
– Werner Vogels, Amazon
The keynote was given to a capacity audience of over 200 people in a Methodist hall in Westminster. The theme that Vogels kept on returning to was that AWS is ready for the enterprise. He stressed that companies from SAP to Oracle have redesigned some of their applications to run in a redundant, high availability format across the AWS cloud. He also encouraged enterprises to get started with the AWS cloud by using the Amazon Virtual Private Network to carve out a portion of their on-premise infrastructure and link it to the AWS cloud.
After Vogels, AWS's head of business development, Dan Powers, took the stage to give even more examples of enterprises using AWS. Shell is a customer and has begun using the cloud for some of its high-performance computing applications, while an American insurer, a Spanish bank, FC Milan and The Guardian are also customers, he said.
He compared the cloud to "like Linux when it first came out," noting that many enterprises were talking about using it, but fewer were talking publicly.
Amazon is working with major software companies around the world — Microsoft, IBM, Oracle, Red Hat, and others — to develop licensing structures for their technology that make it easy to use in the cloud, even on an hourly basis, he said. "I call it bring-your-own-licence," Powers added.
