Industry executives and experts share their predictions for 2021. Read them in this 13th annual VMblog.com series exclusive.
The Software Exodus to the Edge
By Ashwin Mistry, Chief Operating Officer at Taubyte
As of 2019 there were about 27 billion active connected devices on the internet, and growing at a rate of about 127 new devices every second. This translates to over 4 billion new connected devices in 2020. If we extrapolate this over 5 years, this can reach close to 50 billion devices on the plant by 2025. In fact, per International Data Corporation (IDC) estimates, there will be “41.6 billion connected IoT devices … generating 79.4 zettabytes (ZB) of data in 2025”.
Today there are about 3,000,000 Edge data centers in the US alone that are ripe for Edge Computing services; in comparison, hyperscalers generally operate around 20 to 30 “regional” data centers globally. As more devices onboard the internet due to technological advances, especially due to the 5G roll outs, we foresee more Edge data centers penetrating the market. This trend will create the stage for a colossal management and orchestration effort when it comes to administering Edge Computing services using the hyperscalers’ approaches.
Our Predictions
Network Outages Continue for Hyperscalers
Hyperscalers will continue to experience outages due to more connected devices embracing the centralized architectures of Cloud Computing. In a span of about one year, all 3 major Cloud service providers experienced an outage, and caused applications that rely on them to fail. We predict that 2021 will see at least 4 major outages, causing billions of dollars in losses, not to mention the aggravation felt by their users. This ultimately will pave the way for shifting to Edge-native options, like Smart Computing by Taubyte.
Using Cloud-native Tools at the Edge Won’t be a Sustainable Solution
The trend of adopting Cloud-native tools and platforms to the Edge will reveal to be extremely expensive and unscalable. A good example is K3s, a claimed lightweight Kubernetes version for Edge computing. Although those tools seem cool, they require a lot of Ops, and do not make the Edge any more consumable for Developers. The Edge is really a reverse-Cloud, every concept that the Cloud holds dear is broken at the Edge. A good example is the immutability and homogeneity of the Cloud infrastructure. As we see it, using Cloud-native tools at the Edge will require an army of DevOps behind every deployment!
Rise of Smart Computing
With the record number of data centers operational at the Edge, and, as mentioned above, there are about 3,000,000 Edge data centers in the United States alone, the Edge Computing infrastructure is pretty solid, and continues to grow alongside 5G roll outs. This infrastructure is further augmented by the growing number of connected devices. Both those small proximity data centers and the connected devices are susceptible to network fluctuations. This level of complexity requires a resilient solution that Developers can employ to handle not only the hardware variety, but also the myriad of programming languages supported. Smart Computing makes this dynamic Edge infrastructure highly consumable and composable by building on top of the Cloud, and eliminating overhead tasks that are not conducive to the Edge. Hence, we foresee the development community adopting this smarter approach for Edge Computing applications.
Conclusion
Tweeking Cloud-native tools to be used at the Edge is a brute-force method that will reveal to be a cumbersome option in the upcoming two to three years, which is about to experience phenomenal growth, not only in connected devices, but also in proximity data centers. Smart Computing is the only solution that will result in a highly resilient and scalable experience at the Edge. Companies that build their applications on top of a Smart Computing Platform will have an “edge” against their competitors.
##
About the Author
Ashwin Mistry is the Chief Operating Officer at Taubyte, a DeepTech startup based in Dallas, Texas. Prior to his role as COO, Ashwin had taken on various consulting assignments, including serving as the Director of Product Technology at another startup. With over 30 years of experience in Embedded Systems and IoT, Ashwin has the in-depth knowledge and experience in his chosen field to help enterprises make smart business decisions on adopting emerging technologies. His tenure in new product development spans decades among several industries, and gives him an advantage on quickly assimilating recent trends. This has inspired Ashwin to write several blog articles and white papers on the Edge.
