The Emergence of Cloud Computing Platforms for Data Science
Author Dr. Jeong Sihm
Assistant Professor of Math
School of Arts and Sciences
Amazon is one of the most famous multinational technology companies, and many of us are buying all kinds of products and services through Amazon prime subscription. Also, many would be surprised to know that Amazon’s largest source of operating profits is AWS. This subsidiary provides on-demand cloud computing platforms on a pay-as-you-go basis. In personal computers, many organizations once bought computing equipment directly while developing and purchasing their business solutions from the ground up. Nowadays, they purchase convenient, on-demand network access to a shared pool of configurable computing resources such as networks, servers, storage, applications, and services. This business practice has saved organizations a considerable amount of fixed costs associated with owning and maintaining everything independently, allowing them to stay nimble and scale-up effortlessly to meet the ever-changing business demands. Thus, here comes the metaphor of clouds as computing solutions that are no longer installed and managed on-site within business entities. To put it simply, we don’t buy physical computing equipment for on-site ownership anymore. Instead, we pay the fees to access business solutions safely hosted on the best data centers in the world and maintained by Amazon, Google, Microsoft, and the like.
And this strong trend continues in the domain of data science too. Smaller companies now can perform data analytics and run machine learning solutions to compete with more prominent players in the market without worrying about the high costs associated with physical equipment, servers, and facilities. Some notable players in this popular field of Data as a Service (DaaS) include Amazon, as mentioned earlier, Web Services, Google Cloud, IBM Cloud, and Microsoft Azure, to name a few. The popularity of cloud computing platforms for machine learning will undoubtedly change how data science solutions are developed and deployed in practice. Furthermore, it would be a challenge for degree programs like our MS in Data Science to prepare students for this rapid development in Data Science.