You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
With the cost of computational power coming down in cost, size and power consumption more and more computational power will be moved to where it is used the most. It will also require ever increasing specialization of functions. Cost/benefit analysis across entire ecosystems will become the key skill set to those that will win market share. It will be end to end ecosystems and not devices taken in isolation that will win at this game. Ecosystem builders will be the winners.
This trend has been going on already for many years, and is called Network Edge Computing, or just Edge Computing. The primary enabler was supposed to be 5G cellular communications, but for many applications wired networks or 4G was sufficient, and it is not clear that 5G deployment has made the promised difference. This is a decent high-level overview of what's going on in the industry and why:
Learn about edge computing, how it works and the importance of its role in the growth of 5G. Discover why edge computing matters, including benefits and use cases.
www.techtarget.com
My take has been it looks like a security nightmare waiting to happen for some applications, especially in vehicles, but those concerns (which many share) hasn't slowed the trend. The primary motivation for edge computing appears to be the cost and annoyances of cloud computing, like AWS. While cloud computing is cheaper than owning and running an enterprise data center in most cases, it is still expensive, and for some applications that need low latencies to computation edge computing is the only practical answer.
