HPI Issue #9: Cloud Computing and Edge Computing (There is No Versus)
At StackPath we don’t believe there is a versus when it comes to the cloud and edge. We believe that edge computing augments cloud computing. For instance, with IoT use cases, edge computing would perform just-in-time processing and decision making while cloud computing would store large amounts of data and perform big data analysis. In short, the edge and the cloud work together to create a comprehensive solution.
According to one of the latest cloud versus edge articles from Lucas Roh, a computer science PhD and 20+ year cloud computing pioneer (story featured below): “Machine learning and artificial intelligence will all happen in the cloud, but more localized processing will also slowly move to the edge. Thus, edge computing is ideal for applications where every millisecond counts, while cloud computing works best for applications that are not time-sensitive.”
So, while some are speculating that edge computing will overtake cloud computing by 2025, this is from a perspective of market share and not of saying one is better than the other. The demands on the internet are continuously growing and edge computing will make it faster while cloud computing will make it more scalable.
We need both. And whether you’re using edge computing, cloud computing, or both, the shift away from on-prem solutions will require computing to be distributed.
We cover these topics and more in this issue of High Performance Infrastructure!
Cloud Computing Vs Edge Computing (Forbes)
A good read for those curious about the high-level differences between cloud and edge.
- Cloud computing means using remote servers or data centers for storing, processing, computing, and analyzing massive amounts of data.
- Edge computing moves the compute, storage, and networking closer to the source of data, significantly reducing travel time and latency.
- Some analysts claim that edge computing will replace cloud computing because computing will become decentralized and the need for the centralized cloud will fade. But that will not be the case, as their functions are different.
Why Edge Computing Architecture is Over-Hyped (IT Pro)
We don’t agree with this but it’s worth sharing to show both sides of opinions on the topic of edge computing.
- Author claims there’s no metric for measuring what counts as edge computing and what is simply running a cloud workload in a data center that’s closer to certain users.
- Author claims that, in a sense, every good data center architect has always been striving to implement edge computing, long before anyone was calling it that.
- Author is not sure that cutting out milliseconds worth of data transfer delays is really going to revolutionize many end user experiences.
The Future of Computing is Distributed (Anyscale)
Until recently, distributed applications have been the exception rather than the norm.
- There are two major trends fueling this transformation: the end of Moore’s Law and the exploding computational demands of new machine learning applications.
- Under Moore’s Law, processor performance famously doubled every 18 months. Today, however, it grows at a paltry 10-20% over the same period.
- The growing gap between the demands and capabilities of specialized processors leaves system architects no choice but to distribute computations for machine learning applications.
Other Worthy Reads
Edge Computing Reading List (University of Wisconsin-Madison)
A great way for developers to introduce themselves to edge computing. From University of Wisonsin-Madison’s “Edge Computing Laboratory” computer science class.
New Distributed Timeseries Database (Uber)
M3DB, inspired by Gorilla and Cassandra, is a distributed time series database released as open source by Uber Technologies. It can be used for storing real-time metrics at long retention.
Programming Languages for Distributed Systems & Data Management (Schloss Dagstuhl)
This report documents a program for building a community of researchers interested in programming language techniques for distributed systems and distributed data management.
Space Development Agency Evaluates Cloud Computing (Space News)
The SDA is evaluating how cloud computing could bolster its campaign to quickly identify and track missiles.
How to Microservice Your Video Streaming Workflow at the Edge (StackPath)
Use StackPath’s CDN, Edge Compute, and Object Storage with Unified Streaming’s Unified Origin product to deploy a globally distributed video streaming workflow that maximizes performance and minimizes costs.