IDC Research Report | ZStack Selected as a Key Vendor in China’s Generative AI Application Development Platform.

Recently, International Data Corporation (IDC) released the report “China Generative AI Application Development Platform Market: The Emergence of Unified AI Development Platforms for Enterprises” (November 2024), which provides an in-depth analysis of the challenges faced by enterprises when expanding generative AI applications. The report also focuses on evaluating mainstream generative AI application development platform products.

As the industry’s first generative AI application development platform report, IDC’s research found that 2024 is still a phase for vendors to refine their products. Most products are expected to officially launch in the second half of 2024. Generative AI application development platforms/large model development platforms are still in the early stages in the Chinese market, with the vendors included in this report being early entrants. The report outlines the essential capabilities that platforms should have, including data preparation, model tuning support, basic RAG/Prompt capabilities, model deployment, ensuring model and application security, and at least two commercial use cases. ZStack from Cloud Axis Technology was included in the report as a key vendor in the “China Generative AI Application Development Platform” category, showcasing the company’s core strength in enterprise-level AI infrastructure.

Currently, there is no unified market definition for large model platforms/generative AI platforms. IDC, in line with global trends, believes that enterprises will eventually have unified AI development platforms, with large model platforms being an essential function. At this stage, the market is simply defined as “Large Model/Generative AI Platforms.” IDC anticipates that the foundational large model capabilities will continue to improve rapidly in 2024, with generative AI applications beginning to take shape. However, this is still an early stage for the technology, and industry participants should continue to monitor the evolution of foundational models and the latest forms of generative AI applications.

IDC forecasts that by 2028, the Chinese generative AI software market will reach $3.54 billion. According to the vendors included in this report, large model platforms are already showing early development. With the enhancement of large model capabilities and continuous innovation in application forms, it is expected that large model platforms will evolve into two types of products: foundational platforms and intelligent agent development platforms.

The report highlights that in technology investments, the focus is usually on platforms that offer a comprehensive set of functions, rather than standalone solutions. The ZStack AIOS platform, also known as Zhita, developed by Cloud Axis Technology, was included in the report as a next-generation AI infrastructure platform released in August 2024. Leveraging the advantages of its all-in-one platform, it enhances cost-effectiveness. ZStack AIOS is a self-developed, productized, and standardized private AI infrastructure platform that centers on AI enterprise applications. It comprehensively supports AI innovation in various industries from three major directions: “computing power scheduling, AI large model training and deployment, and AI application service development.”

First, the ZStack AIOS platform supports both AI model training and deployment development, assists with AI application construction and service operations, and provides computing infrastructure management. Second, it improves the utilization of heterogeneous hardware, reduces AI costs, accelerates multi-model collaboration, and optimizes AI performance. With full-scale metering and billing, it enables AI self-service, thus accelerating the privatization of enterprise AI applications. Third, by providing pre-configured AI development tools, services, virtualization, cloud-native services, and high-performance storage networks, the platform significantly lowers the barrier to AI application deployment for enterprises. Fourth, in terms of data security, Zhita offers purely localized data management and encryption, high availability self-healing for business failures, and enterprise-grade data disaster recovery services, ensuring comprehensive data security and business continuity.

As the foundational large model capabilities continue to improve and generative AI applications begin to take shape, ZStack will continue to leverage its first-mover advantage in the AI infrastructure platform domain, driving increased enterprise AI penetration and supporting companies in developing more efficient, secure, and innovative AI applications on their digital transformation journey.