Why Software-Defined Everything Is Shaping the Future of Computing

21CTO Community Guide : Combining the current AI boom, Professor Mei Hong believes we are entering a software-defined era. The essence of software-defined technology is to break the original integrated hardware, virtualize the underlying hardware, provide standardized basic functions, and control them via management software for more open, flexible, and intelligent services.

Professor Mei Hong, academician of the Chinese Academy of Sciences, fellow of IEEE, former dean of Peking University School of Information Science and Technology, and current vice‑president of Beijing Institute of Technology, focuses on software engineering and system software, with achievements in component middleware, development methodologies, and tool environments.

His keynote report, "Software Defines Everything: Challenges and Opportunities," is divided into three parts: ubiquitous software, the software-defined era, and the opportunities and challenges of the new era.

He categorizes the development of computer software into three stages:

1946‑1975: Integrated hardware‑software stage.

Post‑1975: Software productization and industrialization stage.

Post‑1995: Software networking and service‑orientation stage.

Combining the AI surge, he asserts that we are entering a software-defined era where hardware is virtualized, basic functions are standardized, and management software provides open, flexible, and intelligent control.

Software technology is one of the core competencies of the new manufacturing revolution. Intelligent manufacturing requires the software‑definition of hardware, knowledge, and processes, ultimately leading to platform‑level software definition.

From an operating‑system perspective, software definition involves hardware resource virtualization and programmable management functions.

Ubiquitous Software

From the infrastructure view, we are in an Internet+ era; from the computing model view, a cloud‑computing era; from the information‑resource view, a big‑data era; and from the application view, an intelligent era. Software is the key technology that underpins all these trends.

Software and information‑technology services have shown sustained high growth in China, outpacing other sectors even during the 2008 financial crisis. Globally, software spending exceeds 0.5% of GDP in 19 countries, with the United States above 1%.

Software productization, driven by companies like Microsoft and Oracle, marked the emergence of an independent software industry, followed by a shift to networked, service‑oriented software with the rise of the Internet and mobile apps.

New Features in the Internet Environment

Chinese scholars coined the term net‑constructed software to describe software paradigms for Internet computing, characterized by autonomy, context awareness, collaborative behavior, online evolution, and trustworthiness.

Software, combined with the Internet, enables true service‑orientation, while mobile Internet introduced the APP model, further changing software delivery.

Software‑Defined Era

The second part discusses the software‑defined era, where human‑machine‑thing integration drives a third wave of intelligence based on deep data mining and fusion.

What Is Software‑Defined?

The practical realization of software definition occurs in cloud platforms. Around 2011, OpenFlow was adopted for network management, and by 2015 Gartner introduced SDN (Software‑Defined Networking), redefining traditional network architecture.

SDN enables programmable control of network devices via APIs, allowing new protocols and topologies without hardware changes.

Operating systems manage hardware resources, provide common services, and embody the culmination of software definition.

Software‑defined technology fundamentally virtualizes hardware resources and makes management functions programmable, leading to software‑defined storage, compute, environments, and data centers, all built on a three‑layer OS architecture.

Opportunities and Challenges of Software‑Defined Everything

Opportunities include achieving universal connectivity and programmability across humans, machines, and things, extending software definition to physical worlds, cities, industries, and campuses.

Challenges span architecture design decisions, system quality (balancing flexibility with performance loss), security risks of programmable hardware management, lightweight virtualization for emerging devices, smooth migration from legacy systems, and building highly adaptive software platforms capable of predicting and managing future hardware changes.

Source: AI Technology Review