Now, companies like IBM are talking about cognitive clouds that can change the nature of computing. Data analysis was always a feature of cloud computing; now, the game is artificial intelligence and blockchain.
In 2017, Marc Clark, then senior director of cloud strategy at Teradata, asked a simple question. “How do you transfer 3PB of data on the cloud?” After a few minutes, he animatedly answered the question himself, “you don’t”. Three years, though, is a long time in computing. Technology has transformed to unimaginable levels. In 2017, companies were still working out kinks on hybrid clouds and moving for better integration between on-site and off-site servers. Now, companies like IBM are talking about cognitive clouds that can change the nature of computing. Data analysis was always a feature of cloud computing; now, the game is artificial intelligence and blockchain. “We are moving from public systems to hybrid multi-clouds with cognitive capabilities,” says Gopal Pingali, global vice president of the Global Technology Services Labs, IBM. But it is crucial to understand what is a private cloud, a public cloud and a hybrid cloud.
What is a private cloud?
The cloud journey started with private clouds, which were installed either on-site or with a third-party operator but were dedicated to solely one enterprise. This offered more security but also constrained the ability of an organisation as it could not scale up immediately. Most companies have long-standing businesses that are run on IT, with sensitive data like transactions carried out on mainframe. “Over time virtualisation has taken place, which has meant that the world is no longer tied to machines,” Pingali explains.
What is a public cloud?
As scalability was an issue that companies could not address, there was a need for services that could be scaled immediately. Meanwhile, the explosion of worldwide web led to the emergence of public clouds, eliminating the need for companies to invest in private clouds or server farms. This provided cost-effectiveness and scalability. The vast network ensured that chances of failure are low. But security with public clouds is an issue. So, every company is either transitioning or has transitioned to the public cloud.
What is a hybrid cloud?
For enterprises that want both scalability and security, the best option is to combine both the services. Pingali says that only 20% or so is in the public cloud. Such a kind of system provides flexibility to companies to keep sensitive data on private clouds and migrate the rest to public clouds. A problem, however, is that clouds need to work effectively across the board. For IBM, Pingali states, the answer has been hybrid multi-cloud management. IBM has been able to use hybrid management systems to merge cloud services to make them more seamless.
What about data analysis?
A cloud system is not just about space or more computing power, and it also entails data analysis. This requires computing power, and this is also where artificial intelligence jumps in. The ability of IBM Watson as a data management system is tested. Watson is expanded to apply to this system. “Since we have the largest enterprise Data Lake in the world. We have the most advanced AI. It can provide deep insights, and present problem areas to focus on and give warnings before an event happens,” Pingali highlights. We have Watson which detects issues automatically. It can also retrieve issues.
What about IoT devices and how do they figure in cloud management?
IoT devices add two more substantial dimensions. Security and compliance with privacy norms are now being tied with hybrid clouds to connect smart IoT devices to drive growth. Autonomous devices can be tied with Watson.
Techsplained @FE features weekly on Mondays. If you wish to send in queries that you want explained, mail us at firstname.lastname@example.org