To receive industry-leading AI updates and exclusive content, sign up for our daily and weekly newsletters. Learn more
Undoubtedly, enterprise data infrastructure continues to change with technological innovation, most notably today due to data- and resource-hungry generative AI.
As Generation AI transforms the enterprise itself, leaders continue to wrestle with the issues of cloud, edge and on-premise: on the one hand, they need near-instant access to data, and on the other, they need to ensure that data is protected.
In the face of this challenge, more and more businesses are seeing hybrid models as the way forward, as they offer different benefits from cloud, edge and on-premise models. 85% of cloud buyers According to IDC, they have either adopted hybrid cloud or are in the process of adopting it.
“The pendulum between edge and cloud and hybrid flavors in between has been shifting over the last decade,” said Priyanka Tenbay, co-founder and CTO of runtime application security company. Operant“We’re seeing a lot of use cases emerge that benefit from running compute closer to the edge or combining edge and cloud in a hybrid fashion,” he told VentureBeat.
The shifting pendulum of data infrastructure
For a long time, cloud has been associated with hyperscale data centers, but that’s no longer the case, explains Dave McCarthy, research vice president and global research leader for Cloud and Edge Services at IDC. “Organizations are realizing that cloud is an operating model that can be deployed anywhere,” he says.
“The cloud has been around for a long time, and it’s time for customers to rethink their architecture,” he said. “This opens the door to new ways to leverage hybrid cloud and edge computing to maximize the value of AI.”
Miguel Leon, senior director at the app modernization company, noted that AI is driving the move to hybrid cloud and edge as models require more and more computing power and access to large datasets. Winwire.
“The combination of hybrid cloud, edge computing, and AI is really changing the technology landscape,” he told VentureBeat. “As AI continues to evolve and become an embedded technology in virtually every business, it will only become more intertwined with hybrid cloud and edge computing.”
The Edge Addresses Problems the Cloud Alone Cannot Solve
IDC research shows Edge spending will grow by 10% $232 billion this yearMcCarthy noted that this growth is due to several factors, each of which addresses a problem that cloud computing alone cannot solve.
One of the most important are latency-sensitive applications. “Latency represents latency, whether it’s introduced by the network or the number of hops between the endpoint and the server,” McCarthy explains. For example, vision-based quality inspection systems used in manufacturing must respond in real time to activity on the production line. “This is a situation where milliseconds matter, and you need a local, edge-based system,” he says.
“Edge computing processes data closer to where it is generated, reducing latency and increasing business agility,” agreed Leong, and also supports AI apps that require fast data processing for tasks such as image recognition and predictive maintenance.
The edge is also useful in environments with limited connectivity, such as Internet of Things (IoT) devices that can be mobile and move in and out of coverage areas or have limited bandwidth, McCarthy noted. In certain cases, such as self-driving cars, AI needs to operate even when the network is unavailable.
Another problem common to all computing environments is the volume of data. Latest EstimatesApproximately 328.77 million terabytes of data are generated every day. By 2025, the amount of data is expected to grow to more than 170 zettabytes, which means a more than 145-fold increase in 15 years.
As data from remote locations continues to grow, so does the cost of sending it to a central data store, McCarthy noted. But with predictive AI, most of the inference data doesn’t need to be stored long-term. “Edge computing systems can determine what data needs to be retained,” he said.
Government regulations and corporate governance can also limit where data can be stored, McCarthy said. As governments move to enact data sovereignty laws, companies will face increased compliance challenges. This can happen when cloud or data center infrastructure is outside of local jurisdiction. Here, too, the edge can help.
As AI initiatives rapidly move from proof-of-concept testing to production deployments, scalability has become another major issue.
“The influx of data can overwhelm core infrastructure,” McCarthy said. He explained that in the early days of the internet, content delivery networks (CDNs) were created to cache content closer to users. “Edge computing will do the same thing with AI,” he said.
Advantages and Applications of Hybrid Models
Of course, different cloud environments offer different benefits. For example, McCarthy noted that auto-scaling to meet peak usage demands is “perfect” for public cloud, while on-premise data centres and private cloud environments help you secure and have better control over your own data. The edge offers on-site resiliency and performance. Each has a role to play in a company’s overall architecture.
“The beauty of hybrid cloud is that you can choose the right tool for the job,” McCarthy said.
He pointed out many use cases for the hybrid model: in financial services, for example, mainframe systems can be integrated with cloud environments so financial institutions can maintain their own data centers for banking operations while leveraging the cloud for web- and mobile-based customer access, while in retail, local systems in stores can continue to process point-of-sale transactions and inventory management independently of the cloud in the event of a disaster.
“This will become even more important as retailers implement AI systems to track customer behavior and prevent inventory loss,” McCarthy said.
Tenbay also noted that sensitive data can be protected through a hybrid approach that combines AI running locally on the device, AI running at the edge, and larger private or public models with strict isolation techniques.
That’s not to say it’s without drawbacks: For example, McCarthy noted that hybrid can increase management complexity, especially in mixed-vendor environments.
“This is one of the reasons why cloud providers are extending their platforms to both on-premise and edge locations,” he said, adding that original equipment manufacturers (OEMs) and independent software vendors (ISVs) are also increasing their integrations with cloud providers.
Interestingly, at the same time, 80% of IDC survey respondents said they have moved or plan to move some of their public cloud resources back on-premise.
“For a while, cloud providers tried to convince customers that on-premise data centres were gone and everything would run in the hyperscale cloud,” McCarthy notes. “But that has proven not to be the case.”
