With the explosion of data, how are high performance computing needs emerging? How does AWS view this opportunity?
Typically, scientists and engineers must wait in long queues to access shared clusters or acquire expensive hardware systems for their high performance computing needs. With the cloud, businesses can roll out hundreds or thousands of servers, even those with HPC power, in a matter of minutes and pay for what they actually use. They are able to store huge amount of data at a very low cost in the cloud. This helps businesses drive costs down significantly while enabling them to analyze enormous amount of data quickly, giving businesses, researchers, analysts and developers the competitive advantage . This is clearly where cloud computing such as AWS can help provide virtually unlimited infrastructure capacity and compute power. There is a tremendous opportunity for AWS by providing businesses, on-demand access to supercomputing power.
Please shed some light on how AWS is facilitating supercomputing needs?
The utility model we use at Amazon Web Services (AWS) is without upfront capital investment, which means that we are able to operate our infrastructure in a very efficient manner, at very large scale for hundreds of thousands of customers across 190 countries.
In high performance computing, we have solutions like Amazon Elastic Compute Cloud (Amazon EC2) that provides several instance types which are designed for data intensive workloads; Cluster Compute instances that offer high performance CPUs and are deployed on a fully bisectional, 10 gigabit network for low latency communication between instances. We also have Amazon Elastic MapReduce (Amazon EMR) that utilizes a hosted Hadoop framework running on the web-scale infrastructure of Amazon EC2 and Amazon Simple Storage Service (Amazon S3). Amazon Redshift is a fast, fully managed, petabyte-scale data warehouse service that makes it simple and cost-effective to efficiently analyze all your data using your existing business intelligence tools. Amazon DynamoDB is a fast, fully managed NoSQL database service that makes it simple and cost-effective to store and retrieve any amount of data, and serve any level of traffic requests.
Today, AWS customers run a variety of HPC applications on these instances including computer aided engineering, molecular modeling, genome analysis, and numerical modeling across many industries including Biopharma, Oil and Gas, Financial Services and Manufacturing. Also, academic researchers are leveraging Amazon EC2 Cluster instances to perform research in physics, chemistry, biology, computer science, and materials science.
As Amazon is very comfortable in operating a high volume, low margin business, and with our high economies of scale and operational efficiency, we are able to drive our costs down and pass those savings back to our customers in the form of lower prices. We have lowered our prices 37 times across many of the AWS Cloud services since 2006. Our customer’s HPC workloads can run at a higher scale, complete their processing much more quickly at a very low cost that they can afford and this opens up new opportunities to experiment and drive innovation. We even provide flexible choices that enable our customers to set the price for their infrastructure costs, saving them even more.
What will be the future of high-performance cloud computing?
At AWS, we work backwards from our customers by listening carefully to what they say is important to them, and this means inventing on behalf of our customers and iterating quickly based on our customer’s feedback. We also focus on what is fundamentally important for all our customers, i.e., continue to drive operational efficiency with our economies of scale to lower the price of infrastructure resources to benefit our customers around the world. We’d like computing resources to be so low cost and ubiquitous that customers could easily have supercomputing power at their fingertips, at any time they need it.