Edge Computing vs. Cloud Computing
There’s been a lot of talk about how businesses should move to the cloud. But recently, a new term has entered the conversation: edge computing. But what is edge computing, and how does it differ from cloud computing? Let’s dive into the differences and similarities in this post.
The Basics: Edge Computing vs. Cloud Computing
Cloud computing is the delivery of computing services, such as applications, databases, and storage, from a remote location, usually provided as an on-demand service. According to Aidan Fitzpatrick, Reincubate CEO, the concept started as a way for businesses to rent out spare capacity from their pre-existing server infrastructure to third parties via the internet.
As the name implies, edge computing involves putting computing resources on the outer edge of a network. The concept emerged from the need to reduce latency in communications, particularly with videos, gaming, and high-frequency trading (e.g., stocks, bonds, currencies), where milliseconds in delays can greatly degrade the user experience. Latency is proportional to geographic distance. The latency will increase if this flow needs to travel a large distance (e.g., between countries or across continents).
To reduce latency, content delivery network (CDN) providers, such as Akamai, Cloudflare, and others (including some that also offer cloud services), established global server networks, re-routing the end user’s traffic so instead of going directly to the server, they were instead directed to the nearest copy of that data on the CDN’s own servers.
Rather than merely caching static data, CDN providers enabled customers to run some of the applications that would otherwise be run on a server or on a cloud computing platform directly on the CDN servers, providing the benefit of low latency without the drawback of storing only static data, Fitzgerald said.
Full Service vs. Microservices
Fitzpatrick said that the big difference in edge computing vs. cloud computing comes down to full service vs. microservices. Edge computing is designed around the concept of microservices, where many small and simple applications handle a specific task and interact with each other. This usually requires that the application is designed for edge computing. Pre-existing applications typically don’t perform well in an edge computing environment.
By contrast, cloud computing enables a company to use the cloud for any computing that would otherwise be done via on-premise servers.
There are notable security differences between edge computing vs. cloud computing.
With edge computing, there’s a risk of losing data, not so much over cybersecurity concerns, but more with damage to the local servers, notes Kevin Gordon, NexOptic vice president of AI Technologies. “They could be prone to natural disasters or manmade errors that cause fires.”
Gordon added that cloud computing is vulnerable to the ever-increasing risk of cybersecurity breaches, which can lead to major data theft or loss.
Additionally, any organization that collects personally identifiable information (PII) is going to have concerns about using a public cloud provider, according to Craig Theriac, Scale Computing’s vice president of product management. “Whether a retailer using digital IP surveillance cameras for loss prevention or a manufacturer using cameras on their factory floor as part of their quality assurance system, storing that type of sensitive data on a public cloud could expose an organization to a raft of liability and/or compliance issues.”
However, cloud computing offers the security of data backup, disaster recovery, and business continuity, which are easier and less expensive because data can be mirrored at multiple redundant sites on the cloud provider’s network, said Łukasz Koczwara, STXNext vice president of engineering.
Public cloud providers also have more resources to draw on to spend on security than individual companies.
Primary Edge Computing Users
According to Craig Theriac, Scale Computing’s vice president of product management, manufacturing, logistics, and retail industries were among the earliest edge computing adopters. They had the most to gain from being able to process and analyze data close to the source.
Businesses that are primarily stateless are the primary users of edge computing, according to Jacob Loveless, Edgemesh CEO. Small data applications like small data (not big data), including APIs, configuration information, and stateless and semi-stateless services, do great on the edge.
Autonomous vehicles must do edge computing — directly on the vehicle itself, Keshani added. “Collision avoidance systems can’t be run in the cloud when your data connection is not reliable, and they need to be low latency.”
Fitzgerald said that edge computing tends to be favored by tech startups that can start fresh and develop their applications and software specifically for an edge computing environment.
Companies Using the Cloud
Small and large companies use the cloud, leveraging the resources of cloud providers such as Amazon Web Services (AWS), Alibaba, Google, Oracle, and others.
“The great advantage of cloud computing is that it allows for tremendous offsite computing, processing, and storage, and nearly ubiquitous global access to these resources so long as we have a network connection,” Gordon said.
One of the key benefits of cloud computing is the ability to access data from anywhere at any time, which allows for more flexibility in regard to where data teams are located and who can access it, said Dan Ciruli, D2iQ vice president of product management. “Cloud computing also allows organizations to scale more easily — as the business grows, the storage requirements that grow along with it are more easily handled through cloud computing. It also provides organizations the opportunity to increase automation so data professionals can spend more time on other tasks that drive the business forward.”
The Advantages of Having Edge and Cloud Computing
Since many businesses need some of the advantages of both, it’s not so much a decision of edge computing vs. cloud computing, but how they can best take advantage of each, said Jonathan Seelig, Ridge Cloud executive chairman. “They will use centralized could infrastructure for computing functions that are well supported by this architecture. They will expand their footprint to edge computing nodes for applications that necessitate a greater, more reliable and more performant infrastructure.”
Seelig added that the reliability and performance achieved with edge computing are a big advantage, but operating an edge infrastructure increases IT complexity.
Edge computing offers simplified management, according to Loveless. However, database offerings are still in their nascent stages.
“We shouldn’t divide the world into cloud compute and edge computing — there’s a continuum between the two,” said Mimi Keshani, Hadean’s head of operations.
It’s important to keep in mind that the edge computing vs. cloud computing decision is not an either/or proposition, and we are now seeing a growing number of use cases in which a combination of both will become common, agreed Theriac.
“Leveraging the public cloud for non-mission critical applications represents an easy way to slash IT costs since you don’t need to manage your own infrastructure,” Theriac said. “However, for businesses that generate significant volumes of data or run mission-critical applications, the cloud simply isn’t an economically viable option due to the high volume of data being generated as well as the latency issues of transporting that data back and forth to a cloud provider.”
Future Consolidation: Edge Computing vs. Cloud Computing
The question of edge computing vs. cloud computing is somewhat moot for the companies that use it now and could be a non-question if the two merge in the future, which Fitzpatrick thinks “is entirely possible.”
“We are already seeing traditional cloud computing providers such as Amazon Amazon Web Services launching edge computing service, such as Lambda, and on the other hand as providers of edge computing put more server capacity into their edge nodes that are increasingly able to offer a wider range of services that can be run on the edge. Ultimately businesses deploy code to the cloud and don’t know or care whether it is being run on an edge node or on a central cloud computing platform.”