This page contains a Flash digital edition of a book.
inside view

Can cloud computing transform science?

Cycle Computing’s Dave Powers tackles the question of the cloud

offerings from Amazon Web Services (AWS) that many enterprises, academic institutions and SMBs use on a daily basis or whether it’s through a more consumer- based cloud service such as iTunes, Netflix or even Gmail, we are all exposed to cloud computing. The point is not that you are accessing hardware and software resources which live in your business or home and which you own; you simply access the hardware or software when you need it, for as long as you need it and then you turn it off and go about your day. Most would agree these consumer


offerings have been wildly successful in fundamentally changing how people interact with technology. When I moved aggressively into cloud computing in 2008 in my role as an engineer in high performance computing (HPC) at Eli Lilly and Company, we were exposed to a whole new world of infrastructure-as-a-service from AWS, the likes of which I had never seen before in an enterprise environment. Things like provisioning an HPC cluster of machines in minutes instead of weeks, OpEx (operational expense) costs that are pennies-per-hour for a machine versus seven-figure capital investments, scalability that was seemingly infinite versus a fixed capacity, granular transparency of costs, low friction to get started, and on and on. We had seemingly opened the wardrobe

of infrastructure and walked through into Narnia. It didn’t take long before we were able to have a proof-of-concept (PoC) HPC environment up in Amazon’s cloud where we would showcase running internal computationally intensive workloads in the cloud on-demand. In the midst of our excitement and enthusiasm of this brave new frontier, the question was posed to me: ‘How will this fundamentally change how we do science at Lilly?’ That was a provocative and profound question that I have not really stopped thinking about since 2008.


y now, most people have been exposed to cloud computing in one form or another. Whether it’s the popular infrastructure-as-a-service

consider include: (1) keeping your data secret by ensuring it is encrypted at all times; (2) disabling and locking down the machine as tightly as possible to minimise opportunities for security breaches machine access, as computers in the cloud are accessible from the internet; and (3) keeping secret keys and account identification credentials safe – do not share them freely. Many of these are common sense tactical

and operational solutions that we should always be doing in our work environments. Many employees have grown a little too relaxed in their internal computing environments at work and don’t worry too much about encryption or access management or security keys along with data that is moved in plain text and in the open from one machine to another freely. That is a behaviour that will certainly have to change with cloud computing if data is to remain protected and confidential. Another issue that comes up frequently,

Cloud computing has essentially

democratised computing and levelled the playing field. Large-scale, or even small- scale, computing is no longer only reserved for those who can afford to invest large capital dollars in infrastructure and people to support it. Because of the cloud and a simple click-of-a-button, scientists now have access to thousands of computer cores at minimal cost. When we level the playing field and open the door of possibilities to a much broader audience, that’s when we see real innovation flourish! If you want proof, just review the

innovation that has come out of Amazon over the past three years ( about-aws/whats-new/). The benefactors of all of this innovation are the scientists and engineers who can harness this innovation to drive new ideas, more complex questions, and new avenues of research that simply were not available until cloud computing opened this new world of computing. Despite great enthusiasm for cloud

computing there are still some things to be wary of when it comes to the end-user scientists and engineers. The number one issue is security. There are fundamentals that need to be considered when using cloud computing resources securely. Without going into geeky detail or creating an exhaustive list of things to do, there are particular issues when moving into cloud computing for science and engineering workflows. Things to

especially with computational scientists and engineers, is the idea of moving large amounts of data into and out of the cloud. Care should be taken to consider volume of data, throughput, latencies and workflow dependencies when considering moving data into and out of the cloud. For example, you would not want to have 1,000 analysis jobs running in the cloud that need to access data that is back in your own data centre or vice versa. Likewise, you don’t want to have your

workflow running in the cloud and then pause for long periods of time while it waits to upload another large chunk of data from your internal environment. There are several technologies and solutions to help work through some of these issues. Data transfer and data placement technologies and architecture are catching up with large data demands. Ultimately, the network bandwidth available between your facility and the cloud provider are going to be the primary indicator of what your capacity and eventual solution architecture and implementation will look like. It is an exciting time for cloud computing

enthusiasts and innovators. As more and more people become familiar with and comfortable with using cloud computing as the platform for their every-day work, we will see a whole new wave of innovation. This will facilitate visible transformation in how scientists and engineers think about their problem sets and how they attack them and this will result in bold new approaches to how they go about the business of science and engineering.

Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48