by Kayla Matthews, tech journalist and writer
Edge computing is a new form of data processing quickly taking over the jobs of the cloud. More aptly, processing on the edge does all the work the cloud could never hope to do. Edge computing is faster, more secure and more powerful than processing data either in the cloud or in an entirely different space.
The edge refers to the edge of the cloud, right where a device meets the wider world.
Instead of traveling all the way into the cloud to process information, data can be looked into right on the edge, right next to the device, to save on lots of things that are usually hurtful or time-consuming.
While this all sounds like a great plan, companies aren’t latching on to edge computing right this second for a reason.
We still have to resolve some serious problems before we can fully utilize the edge — here are six of them.
1. Power Supply
In order to work optimally, the edge has to be able to process anywhere. Unfortunately, every place in the world may not have the same sort of power supply to run the servers needed for this job.
Making the servers universally connected sounds like an easy fix but can be a lot harder when putting the idea into practice.
Regardless, some machines are becoming able to take power to a power shelf and distribute appropriately. This just has to become a less unique and more common function.
The actual spot houses two problems — available space and physical environment. Some areas, like a college campus, will need a lot more processors than, say, a family farm. All of those processors need servers, and those servers need actual physical room to exist.
Then comes environments where these machines can’t work optimally in wet, warm or humid areas. When thinking about where to put these servers, you must consider everything.
When talking about actual hardware, we already know maintenance will be involved at some point in time.
Maintenance causes problems when people take more time to fix something than a machine could. Any amount of time a server goes down is very bad for edge computing, so allowing the device to already know when servers are going down and fix them automatically is something a few companies are already working on, completely cutting out the human aspect.
Having new, working servers constantly on standby will take more space but is the ultimate solution for now.
4. Artificial Intelligence
Part of knowing when parts will break down in advance has to do with a lower form of artificial intelligence (AI), but this aspect does more than just optimize hardware. For data to be trafficked accordingly, the AI must be able to judge what information is important and what isn’t.
The AI needed to manage data and operate repairs is what will make servers all run at once with affordability in mind. AI will make for faster processing on the edge and subsequently in the cloud and with devices.
While processing data on the edge is safer for everyone involved, this technique isn’t foolproof from malware, just like anything else.
The problem here is if the data on the edge is hacked, then all the devices working on the data are also compromised, making for a much more significant risk than previously expected. For edge computing to become safely operational, the security measures will need a massive overhaul, which could take time.
5G networks are the future of internet browsing and data processing, and edge computing is a crucial component of making sure 5G can work seamlessly. The problem is that edge computing needs 5G to function optimally in the first place.
With faster speeds, edge computing can communicate with the Internet of Things (IoT) and account for latency with devices all at once, leaving room in the cloud for better processing and networking.
Skimming the Edge
We’re literally on the edge of edge computing if we can just hurdle over a few of these gaps. Some companies like Intel have been working on edge computing for years and are thinking of new ways to get around these problems.
A server structure called Open19 is already chugging through these problems to make everything work better.
The Open19 has already fixed the issue of power and is currently working on how to self-evaluate for basic repairs without human interference.
So while the problems listed here are real issues to breakthrough for a new realm of technology, they’re not impossible.
The realm of edge computing in an everyday capacity could become normal before we even realize there had been a global shift at all.