There is a constant buzz in the industry – is Cloud Computing dead? Cloud spend is getting taken over by Serverless and Edge computing. Here’s what that means for the next generation of startups.
Our company, like many of its fellow startups, began life on the cloud. The dotcoms of the previous generation dealt with the economics of managing and scaling their own servers, networks and data centers. But, as luck would have it, we grew up in the pay-for-what-you-need and scale-up-on-demand, “cloud-native” universe.
We’re no members of the Psychic Friends Network, but we’ve seen the cloud’s future. It lives on the edge. This solution is faster. It’s more scalable. And it’s dramatically cheaper.
Recommended: Securing the API Economy In 2022
Here’s why.
First, let’s back up and talk about what edge computing is.
What is Edge Computing?
Edge computing is the ability to run code at the network edge—specifically the content delivery network providers. Content Delivery Networks (CDNs) have been around since the internet has existed. The major CDNs (Fastly, Cloudflare, Akamai, Limelight, Verizon Edgecast) ensure content is delivered quickly to consumers by ensuring a large distributed global server cache.
These providers solely stored data for companies, ensuring that as customers visited websites or downloaded software, there were fast response times because the server itself had the content as close to the customer as possible.
Recommended IT and Cloud Technology Insights 2022: ITechnology Interview with Dominik Pilat, Field CTO at Hivecell
In the past two years, we’ve pretty much abandoned the cloud-computing providers. We now run 95%+ on edge-enabled and serverless infrastructures. So, we have no servers to manage, no locations to spin up.
Even better, there are no cloud computing contracts to analyze.
Our API performance has been not only globally stable but also five times faster as a result. Our infrastructure support has decreased by a factor of ten. The best part?
Our infrastructure spending has decreased by 72% annually.
The future of computing for an entire generation of companies will be “edge-native,” and the traditional cloud is the one that will lose.
There are two changes that differentiate Edge from Cloud Computing.
- Server resources and the CDN are now programmable. This allows companies to move our core API services off centralized cloud servers—and onto the existing globally distributed networks operated by the CDNs.
- The second major change: how the code itself is deployed. For cloud computing, you rent a server and run your code on it. For edge computing, you just deploy your code to the platform. Then, the code runs automatically on the nearest server.
The Economics of Edge
The economic advantage of cloud computing, when it first came to market, was instantly obvious. It allows users to swap high upfront investments in their own server and network infrastructures for zero upfront—and leases on-demand of compute power that can be charged to a credit card. The economics drove the adoption, not the technology.
As for edge computing, there’s similar economic staying power.
Cloud-computing companies buy server and data center infrastructure in bulk to support resales as capacity. But edge platforms, like CDNs, use edge computing to drive additional value on existing infrastructure. This, in turn, actually lowers the cost required to provide computing services to customers.
Top-read Edge Computing News: Veea and Conxai Fuse Edge Computing, Connectivity, and AI Technology
Let’s look at CDNs.
Fundamentally, they’re simple servers, in that they hold copies of data (storage/memory). They look up requests for this data (CPU). Then, they return that data to the user (network). With serverless models, CDN companies are able to further monetize their existing capacity. This allows for meaningful economic impacts down the line.
For example, the closest comparative “cloud-native” and “edge native” offerings are AWS Lambda@Edge from Amazon and Workers from Cloudflare. Let’s compare the two. For AWS: charges are incurred for the request itself, the compute time used by the running code and the network bandwidth used by the response.
For Cloudflare Workers: the cost is a flat fee per request.
So, 100 million requests with a 10 kilobyte response time in the U.S. would cost:
- AWS Lambda@Edge: $95.63 ($75.63 for compute + $20 for bandwidth)
- Cloudflare Workers: $50 ($5 base fee + 90* $.5 for over 10M requests)
The cloud-based bandwidth costs can increase by ~6x as you extend globally. But they aren’t a factor when it comes to edge-native pricing. Edgemesh operates ~2 billion requests per month all around the world.—meaning, our cost savings work out to 72% monthly for “edge native” as opposed to “cloud-native.”
That figure is in addition to the way lower management costs—meaning, no servers to monitor, rapid global scale (code runs near users automatically), and simplified billing.
Startups are offering global, low-cost solutions. The economics—and the simplicity—of the edge-native model is a compelling one. We envision many new startups adopting a cloud-free model as they come of age.
Why wouldn’t they?
There are no servers, no regions to pay for. There are simply on-demand functions running and scaling worldwide. And they can be paid with a credit card.
[To share your insights with us, please write to sghosh@martechseries.com]
