Embracing Long-Termism in the Age of AI

Advertisements

As we venture further into the era of artificial intelligence (AI), the implications of this technology are vast and profoundThe rapid advancements in large models and AI technologies have created an insatiable demand across various sectors for new forms of productive capabilitiesHowever, these advancements are also instituting significant changes within the Information and Communication Technology (ICT) landscapeIn light of this ongoing technological revolution, the restructuring of the ICT industry ecosystem appears to be an unavoidable trend.

But what are the notable influences of AI on the ICT industry? How can ICT harness the momentum of AI to empower various sectors effectively? And what kind of new ecosystem does the ICT industry require in this AI-driven age?

Recently, ZTE Corporation unveiled its strategic insights at the 2024 Cloud-Network Summit, aiming to address these pertinent questions.

In recent years, the accelerated integration of the digital economy with the physical economy has led to a consensus across industries about the necessity of creating new forms of productive capabilities

AI is playing an increasingly pivotal role in this transition.

At the outset of the year, stakeholders emphasized the importance of deepening research and applications related to big data and AI, launching an extensive “AI+” initiativeEssentially, “AI+” is about integrating AI into existing productivity frameworks to drive substantial gains in total factor productivity, ultimately leading to high-quality development.

Even though large models and AI technology are still in their infancy, their impact on infrastructure, application scenarios, and the industrial ecosystem is already profoundTo genuinely realize the potential of “AI+”, it is crucial for market leaders to engage both supply and demand sides, linking the entire industrial chain to construct an open and sustainably developing new ecosystem.

For example, the demands for AI training and inference are rapidly reshaping corporate infrastructure

Innovations in intelligent computing and AI clusters are diversifying computational power, positioning AI-centric infrastructure as a key focus for many companies’ digital transformation effortsAccording to forecasts from IDC, by 2028, AI servers will comprise 33.6% of global server expenditure, with a compound annual growth rate of 17% over the next five years.

Moreover, the integration of “AI+” enhances business scenarios and offers broad avenues for expansionHowever, no singular large model can fulfill all business scenario requirements; choosing the right model for the right scenario will be the most critical pathway to implementing “AI+” effectively in the futureAdditionally, given the rapid evolution of large models and other AI technologies, businesses must leverage the ecosystem to ensure various models are adopted and utilized effectively.

Currently, the landscape of large models and AI technology is still developing, featuring a myriad of technological frameworks, yet the industrial ecology remains fragmented, with significant disconnection along the supply chain

There is an urgent need for market leaders to take the initiative, driving innovation and application to accelerate the restructuring of the ICT industry and hasten the evolution of a new ecosystem.

ZTE Corporation is committed to constructing an open and empowering ecosystemActing as a “builder of the digital economy,” ZTE’s approach involves strengthening foundational technological innovations, promoting collaborative development across the supply chain, enhancing new types of infrastructure, and deepening industry application empowermentIt aims to co-create strategic emerging production capacities alongside partners while enhancing the productivity of traditional industries.

During the 2024 Cloud-Network Summit, ZTE emphasized its goal of establishing the best intellectual network foundation—open to empowerment, tackling the more complex challenges while leaving simpler tasks to its clients and partners, thereby co-building a new cloud-network ecosystem.

But what constitutes the most challenging aspects within this industrial ecosystem? What strategies will ZTE implement to simplify processes for its partners?

For ZTE, a foundational component for the integration of “AI+” lies in infrastructure development.

One of the main challenges facing infrastructure today is how to fully unleash the potential of AI computing power

alefox

The rapid advancement of large models has rendered them “power consumers,” as highlighted by OpenAI's AI Computing Power Report, which reveals the demand for AI computing power is increasing at a staggering annual rate of ten timesConversely, the utilization rate of AI clusters—critical for training and inference—is low, indicating a significant underutilization of computing resources.

As AI clusters scale up rapidly, reaching tens of thousands to hundreds of thousands of units, the complexity of managing these clusters also increases, raising the question of how to capture the full value of computing power.

The Chief Development Officer at ZTE, Cui Li, identifies two key issues that need addressing: first, the application of new technologies to reduce the cost of computing resources, making them affordable for businesses; second, creating a proactive AI cloud-native environment that decouples industry applications from various intelligent computing architectures, allowing for single-time development that can be applied broadly, thus ensuring businesses can leverage computing power effectively.

To this end, ZTE proposes a holistic approach that encompasses hardware capabilities, software capabilities, engineering capabilities, and industrial ecosystem collaboration through a comprehensive AI computing solution

Currently, ZTE remains one of the few providers in the market capable of offering end-to-end, full-stack AI computing solutions, boasting a complete range of products covering various scenarios and an efficient and user-friendly AI platform for addressing the engineering challenges associated with AI training and inference.

In reality, ZTE has established substantial expertise across the fields of ICT communication, networking, and computing, leading to the development of a competitive full-stack AI computing solutionAt this Cloud-Network Summit, ZTE reaffirmed its commitment to ramping up research and development investment in computing power, facilitating a shift from mere connectivity to a paradigm of “connectivity + computing power” to excel in its challenging endeavors.

Apart from infrastructure, the implementation and innovation of AI across vertical industries also hinge on the effective application of various models

Vertical sectors typically consist of specific data, knowledge, contextual circumstances, and business logic that create rich yet complex scenariosThis often necessitates the integration of diverse models alongside predominant vertical large models, indicating that the task is intricately complex.

For this reason, ZTE has devised a strategy known as “1+N+X” for its large modelsIts Nebula large model incorporates both self-developed and ecosystem partnerships, pre-training numerous domain-specific large models to derive various applications, thus accelerating digital transformation across vertical sectorsTo facilitate diverse industries’ utilization of AI for innovation, ZTE also debuted its Digital Nebula 3.0 solution at the summit.

“Digital Nebula 3.0 resolves challenges associated with large models in data, training and inference, applications, security, and deployment

This solution effectively aids enterprises in rapidly creating large model applications while allowing ecosystem partners to continuously participate and collaboratively explore scenarios for AI applications across industries.” According to Hu Xuemei, Vice President of ZTE and General Manager of the MKT and Solutions Division for enterprise clients, “Clients shouldn't focus on how many large models exist but rather on how to meet the demands of their business scenarios.” Digital Nebula 3.0 is designed to satisfy users' innovation needs in fragmented and diverse scenarios with an architecture that is flexible, agile, and capable of rapid iteration, aiding users in their response to swift changes in AI technology and market trends.

In summary, ZTE is leveraging its advantages in core technological innovation, network-computing synergy, and cloud-edge-end collaboration to cultivate AI technology as a foundational capability

Moreover, the integration of AI across product planning and development is driving comprehensive upgrades on product and solution levels to support governmental and corporate users in their digital and intelligent transformation amid the AI age.

It is undeniable that the intelligent transformation sparked by large models remains in its early stages, with considerable ground to cover in both foundational technology and commercial application scenarios.

At this juncture, while the excitement surrounding AI investments is palpable, establishing a healthy and productive industrial ecosystem will be crucial for sustaining the momentum of digital transformation across various sectors in the futureThis endeavor calls for industry participants to embrace a long-term perspective, gradually advancing ecosystem development with an open-minded approach.

ZTE is acutely aware of the enduring nature of ecosystem construction

After years of persistent efforts, the company has seen both the quantity and quality of its partnerships steadily improve, collaborating with over a thousand leading industry partners to create more than a hundred scenario-based solutions, thus establishing a robust cloud-network ecosystem.

Furthermore, ZTE will work on three major fronts to promote the flourishing of the cloud-network ecosystem: optimizing the product ecosystem, broadening the industry ecosystem, and solidifying the commercial ecosystem.

“Optimizing the product ecosystem” involves ZTE engaging upstream and downstream partners while concentrating on the most challenging aspects of foundational technology to create the best possible scenario-based solutions in terms of products and applications“Broadening the industry ecosystem” leverages the “Digital Nebula” platform to increase the number of solution-based and coach-like partners joining the cloud-network ecosystem, thereby exploring the best pathways for the digital transformation of industries

Meanwhile, “solidifying the commercial ecosystem” centers on provincial representatives safeguarding their rights and interests and facilitating collaboration to expand commercial markets across provinces.

“ZTE will operate with a 'large enterprise, small teams' model, allowing for more efficient allocation of company resources and enhanced responsiveness to industry client demandsThis operational approach will integrate both ecosystem and capabilities externally to better unleash productive potential, while internally harmonizing the company’s research and development capabilities, product quality standards, and brand influenceUtilizing the agility possessed by smaller teams will enable ZTE to address the dynamic nature of the government and enterprise markets, respond swiftly to client requirements, and tailor solutions accordingly, leaving simplicity to clients and partners while navigating complexities internally to enhance client satisfaction and create greater developmental opportunities,” stated Xie Junshi.

Overall, ZTE possesses profound insights into the current upsurge in AI technology, with a comprehensive strategy embracing AI in product technology that fortifies its competitive stature