State of the Cloud: Where We Are and Where We’re Heading
Today’s organizations stand at a crossroads with their cloud computing strategy and their investment in cloud infrastructure and tools. Should they continue with their current architecture, or is now the right time to adopt a different approach that better meets their needs and the needs of their customers? Increasingly, companies are choosing to move to a distributed and more open approach to the cloud by shifting from a platform-centric approach to a cloud-native one.
In the recent State of Cloud 2024 webinar, two Akamai customers joined me in a wide-ranging discussion, which included debating the drivers, benefits, and challenges of moving to a more distributed cloud computing model. In other words, a model that decentralizes cloud resources and services so that they are physically closer to the data source or user, while still being managed centrally.
My two guests were Mike Kasprzak, the co-founder of Ludum Dare, and Dolo Miah, CTO of Linebreak. Ludum Dare is recognized as the world’s largest online game jam event, with each weekend event resulting in thousands of brand-new games created by gaming professionals and hobbyists. Linebreak is a global edge-first solution provider that unlocks efficiencies for enterprises in the post-digital era through architecture, solutions, and engineering.
Throughout our conversation, we discussed different cloud computing models and how organizations are using them. I invite you to watch the full webinar. In the meantime, this blog post presents some of the highlights.
Drivers to go distributed: Customer experience, data sovereignty, and cost savings
“Lowering friction is really important to allow our players to play the game right away,” Mike told us, as he shared more about Ludum Dare’s use of cloud computing. Improving customer experience continues to be a strong driver to a distributed cloud computing model, whether it’s delivering instant access, improved responsiveness, or better interactivity.
At the same time, Dolo noted that the shifting compliance environment is another major factor. “Nations are concerned about data sovereignty and understand the value of data as an asset,” he said. “If you’re a global enterprise, data governance and data compliance is absolutely a pressure point that you’re thinking about.”
Organizations are eager not to move data outside of legislative zones if they don’t need to do so. As Ludum Dare scales up its events, Mike too has had to rethink how to meet the needs of changing data compliance guidelines around the world.
Among our customers, I’ve also seen a large move toward cloud cost optimization, which has led to an examination of existing architectures. Starting during the COVID-19 pandemic, this shift to distributed cloud computing continues as more organizations are figuring out how to use multiple clouds, how to take advantage of specific capabilities of individual clouds, and then how to integrate them together.
The benefits of going distributed
The organizations that are quickly adopting a distributed cloud approach are those in industries that have urgent needs for low latency such as gaming, media, and retail. These are the customers with large bandwidth considerations, where they don’t necessarily want to move large amounts of data back to a centralized server and want to be able to act on that data where it’s collected. Artificial intelligence inference is a good cross-industry example of a technology that can deliver the most value when it’s as close to customers as possible.
The amount of data being generated by smart sensors is already vast. It’s not economically feasible for organizations to try to transmit all that data back to a centralized cloud, since they’ll be paying for data storage, as well as ingress and egress bandwidth. “The latency also absolutely doesn't make sense from the point of view of being able to monitor and then react to physical events happening thousands of miles away,” Dolo said.
Enterprises need to become more efficient and ethical with the resources that they have. They must avoid introducing unnecessary costs and waste as well as potentially opening up a broader attack surface.
Going distributed requires a change in thinking
With all these potential benefits, why are some organizations still hesitant to adopt a distributed cloud approach? I agree with Mike that it can be “really hard to rethink your approach” to factor in new paradigms and to constantly re-examine what’s possible. Organizations need to look at the types of cloud architecture that are available, potential partners, and the services they can use to help rearchitect and distribute services globally.
Dolo also talked about the “barrier of thinking differently,” moving from a huge, scalable central cloud to having distributed areas of compute, storage, and networking, each with their own different profiles of how you can scale it along with the need for data synchronization. “The answers are there,” he said, in new architectures, additional skills, and in the awareness that there is an emerging, viable alternative to centralized cloud.
As new technologies and their open source alternatives mature and become available, organizations want and need the freedom to take advantage of those options. However, if they’re hardwired into a particular cloud provider’s architecture and managed services, they may find it much harder to unpick that lock-in and move easily.
What’s next on the cloud journey?
The best way to describe any organization’s cloud strategy, including our own at Akamai, is as an ongoing journey. Since we acquired Linode in 2022, we’ve doubled our cloud computing capacity and are integrating and rearchitecting our platform to enable our customers and ourselves to move workloads to the edge and back to centralized locations.
What we want to create is essentially a breathing cloud infrastructure, which scales to where customers are, so it “wakes up” with our global customers and “goes to sleep” when they do. It’s all about providing more control and automation to our customers so they can run their workloads where they choose, making decisions based on latency, cost, and sustainability.
Future goal: More computing at the edge
For many companies, the first stage of their cloud travels was to invest in and adopt a centralized cloud; now, they’re moving to a distributed infrastructure, with a future goal of conducting more computing at the edge, effectively bringing cloud to the edge.
“Bringing cloud to the edge is the kind of answer to what we’re looking for — to be able to do something a bit more dynamic, closer to my customers,” Mike said. Having one deployment that’s occurring everywhere would remove a lot of the manual effort involved in managing many extra servers. “It allows me to compete with larger companies, if I don’t have to manage so much infrastructure in little pieces all over the place,” he added.
“Part of the reason why we moved across to Akamai is because we see that they're on the same journey as we are with our enterprises — the edge and distributing computing as a whole new way of unlocking value and leveraging data,” Dolo said. Here in early 2024, the focus is on taking first steps, with the journey set to become easier as more organizations embark on it.
Further watching and reading
Watch the entire State of Cloud 2024 webinar
In the opening remarks during the webinar, we referred to recent research on distributed cloud — read the full report: Distributed Cloud: Technology’s Next Act