TECHNOLOGY TREND
What is Next After Cloud 1.0
In the past decade, cloud computing has been a key part of almost all major enterprises' IT strategy. Mid-sized and small businesses have also been leveraging cloud computing to help them bootstrap and scale their IT infrastructure without large upfront capital investment in servers and network equipment. Although there used to be concerns around security and migration overheads, as public clouds keep evolving their technology stacks and improving their operation capability, many enterprises start to consider the Cloud safer, more scalable, and integral to agile IT practice.
In 2020, as the global COVID-19 pandemic pushed companies over the technology tipping point and sped the adoption of digital technologies by several years, running in the cloud is no longer just an addition to enterprises’ IT strategy. Instead, it becomes the core of their strategy for building a future-proof global digital infrastructure. And the benefits that enterprises are looking for from the cloud are not just around cost savings, locations, and instant scalability, but more about the speed of innovation and business agility. To serve the enterprises, public cloud providers such as AWS, GCP, Microsoft Azure, AliCloud all have the foundational digital infrastructure building blocks in place and keep adding new ones to expand their portfolios. In the meantime, they are also going more in-depth into vertical integration to deliver more direct business impacts to enterprises.
However, cloud computing cannot be the silver bullet for solving all the challenges that enterprises have encountered for building the next-gen future-proof global digital infrastructure for scaling their businesses. And there will be new challenges that enterprises have never faced in the past due to ever-changing customer behaviors and preferences and the appearance of new applications and use cases.
Regardless of how much benefits cloud computing provide, one constant debate about the viability of cloud computing is that some applications and data cannot be easily be migrated and operated in the Cloud due to various reasons such as compliance regulation, performance, data volume, security, etc. Often asked questions to cloud architects include:
- Is all workload cloud-native?
- Should all the workload be processed in the Cloud?
- How about AI/ML applications that take lots of data and compute resources to train a model, but the trained model often needs to be applied to real-time generated data at edge locations for deriving actionable intelligence right away?
- Do we need alternative architecture to cloud computing to accommodate what can’t or/and shouldn’t be moved into the Cloud?
- How are 5G and the proliferation of IoT technology going to shape the new architecture?
- How is security going to be handled?
Similar questions also got asked during Future Compute 2021 conference organized by MIT Technology Review last week. This series of blogs are targeting addressing those questions and provide food for thought. Here is a brief overview of what will be covered in this series:
- Both edge and edge computing are very loosely defined terms in the industry. In Part I: what is edge and edge computing, I will discuss some fundamental related concepts in the domain and demystify edge computing and distributed edge clouds.
- In Part II: Why to Care about Edge Computing, I will share my perspectives on imperatives that are driving workload to the Edge and why it makes sense for enterprises to bring compute capability closer to the Edge where data originates, instead of getting data to the compute capability sitting in the Cloud.
- As we start to place a large amount of networked compute and storage capability at the Edge, more workload might be originated at the Edge or be moved from the Cloud to the Edge, which can, in turn, further push up the demand for networked compute and storage capability at the Edge. With the high-bandwidth, low-latency ubiquitous network connectivity promised by 5G, this trend will be further accelerated. In Part III: Edge Computing Use Cases, I will share some edge use cases and discuss some unique challenges to data and AI/ML in Part IV: Data at the Edge: Opportunities and Challenges, and security in Part V: Securing the Edge of a Global Digital Infrastructure.
- Addressing the unique challenges in edge computing requires us to do lots of rethinking around how applications should be designed and deployed and, consequently, how we orchestrate and manage compute, storage, and network altogether. I will share some initial thoughts on this topic in Part VI: Rethinking Infrastructure Orchestration and Manageability in Edge's Era.
It will likely take me a few weeks to finish all the topics, and I maybe add a few more as I discover them while writing up those topics. Please be patient with me and feel free to comment and share your thoughts.
