In this special guest feature, Dr. Viral B. Shah, CEO of Julia Computing, offers 3 Julia language-based examples of the way that cloud computing is revolutionizing data science. Dr. Viral B. Shah is one of the creators of the Julia language and co-founder and CEO of Julia Computing. Julia combines the ease of use of Python with the speed of C. It has been downloaded over 30 million times, and is now taught at MIT, Berkeley, Stanford, and many universities worldwide. Dr. Shah and two other Julia co-creators were awarded the prestigious James H. Wilkinson Prize for Numerical Software in 2018. Dr. Shah earned his PhD in Computer Science from the University of California, Santa Barbara.
In recent years, improvements in cloud computing ease of use, speed and capacity and lower costs have resulted in increased use of cloud computing to solve important problems, including analyzing the spread of COVID-19, pharmaceutical safety and school bus routing.
For example, this dashboard, built using Julia, demonstrates how open source data, in this case, from the New York Times, can be presented for data scientists and other users to inform everything from personal decision-making (e.g. how safe is it to travel to a particular location during the holidays) to public policy.
This free dashboard is hosted in the cloud, and all of the code is available for users to analyze, copy, modify and use.
Another area in which cloud computing is revolutionizing health research is pharmaceutical safety. Pumas-AI uses Julia in the cloud for every stage of pharmaceutical development and testing. The cloud makes this possible by providing access to infinite and immediate scalability for models involving millions of data points and complex simulations.
A third area in which cloud computing is having a real-world impact is school bus routing. AlphaRoute uses Julia in the cloud to design optimal school bus routes for school districts including Boston and San Francisco. The superior computational capabilities of cloud-based infrastructure power solutions that save school districts tens of millions of dollars, allow students to follow sleep schedules that are age appropriate and help them start school rested and ready to learn, reduce traffic and carbon emissions.
These are just three examples of the way that cloud computing is revolutionizing data science.
Why is cloud computing so powerful?
- Scalability: A single user operating a laptop or desktop computer can instantly scale up to dozens, hundreds or thousands of processing units.
- Cost: With pay-as-you-go pricing, you only pay for the computing power you actually use. No need to invest thousands of dollars up front in hardware.
- Access to Latest Technology: The newest, fastest and best processing units – including GPUs, TPUs and more – are instantly accessible online.
- Flexibility: No long term commitment to any particular number of compute hours, number of processing units, hardware configuration or type of hardware. Instantly scale up or down, shift from one type of processing unit to another.
- Democratization: The cloud puts the power of a supercomputer or computing cluster in the hands of any individual user.
- Speed: With instant access to so much top-of-the-line hardware at the click of a mouse, users can run complex simulations and calculations faster than ever.
- Power: The cloud allows users to crunch more numbers and run complex models and simulations that would have been otherwise impossible.
- Openness: Computing power in the cloud can be accessed using any major computing or data science language and any major data format.
- Ease of Use: Any user can easily access the cloud – no need for an IT professional to set up, access or use new hardware.
- Maintenance: Each network is managed by its operators. Users do not need an IT professional or department to maintain an expensive cluster or server farm.
Today’s users are looking for opportunities to reduce cost, analyze massive quantities of data, increase the speed and power of their calculations, access the latest and best new technology, put greater computing power in the hands of their data scientists and analysts and outsource the building and maintenance of expensive hardware. Cloud computing allows organizations to access more and better computing power more easily and at lower cost.
Sign up for the free insideBIGDATA newsletter.
Join us on Twitter: @InsideBigData1 – https://twitter.com/InsideBigData1