Who is the King of Computing Power in A-shares?
Advertisements
The world's ever-accelerating shift towards digitalization has placed a spotlight on computational power, particularly in China, where the demand for intelligent computation is surgingRecent reports indicate that in 2024, China’s intelligent computational capabilities are expected to reach an astounding 725.3 exaflops, marking a year-on-year growth of 74.1%. The market size is also projected to hit $19 billion, an impressive increase of 86.9% compared to the previous yearThis remarkable growth trajectory highlights the increasing importance of computational resources as drivers of economic and technological advancements, especially in the field of artificial intelligence (AI).
The recent surge in interest around AI has been fueled by several high-profile eventsAmong these, the launch of the DeepSeek-R1 model captured significant attention globallyThis new model employs optimized algorithms that promise to enhance the efficiency of computational resource utilization, leading to debates about the potential returns on investment in computational powerAdditionally, on February 17, xAI's Grok3 model was introduced, utilizing a vast network of 200,000 NVIDIA GPUs to push the boundaries of AI capabilities, with Elon Musk claiming it outperforms all existing models.
As the cost of entering the AI industry diminishes, analysts suggest that this will not reduce demand as one might expect; rather, it will likely boost total demand overallDeepSeek, which operates exclusively on open-source models, empowers a wide array of developers with access to advanced AI capabilities, enabling them to create, deploy, and refine AI solutions more flexiblyThis situation echoes the "Jevons Paradox," which posits that as technological advancements improve efficiency, total consumption of the resource may actually increase due to lower costs spurring higher demand.
Meta Platforms' founder, Mark Zuckerberg, recently articulated the strategic value of investing heavily in capital expenditure and infrastructure
Advertisements
He asserts that companies capable of establishing robust infrastructures will gain a significant advantage in delivering quality services at scaleSuch insights reflect a growing recognition among industry leaders that substantial investments in computational capabilities will yield long-term benefits amidst the competitive landscape of AI.
Investment in the computational sector has seen dramatic spikes as market participants recognize the boundless potential of the AI landscapeThe computational leasing index surged nearly 31% soon after the New Year, placing it atop the leaderboard of concept sectorsParallel to this, other indices, such as the “East Data West Computation” index, indicated a rise of almost 24%. Stocks involved in computational services, such as "Bingheng Technology," have soared, with some companies recording explosive gains surpassing 200% since the beginning of the month.
A significant question remains for potential investors: Which players will successfully navigate this AI boom and transform their concepts into tangible profits? Companies poised at the intersection of AI chip development and extensive computational resources, along with those boasting robust data center operations, are anticipated to emerge as frontrunners in capitalizing on this growth.
Turning to key players in the AI chip segment, we find Cambricon Technologies, a leading domestic producerTheir chips are integral to a spectrum of applications including cloud computing and edge computing, thereby tapping into the burgeoning market demandSimilarly, Sugon Information Industry, a prominent IT service provider, is recognized for its high-performance computing products, which enjoy a strong reputation in both domestic and international markets.
Another noteworthy company is Inspur Electronic Information Industry Co., which has established itself as a preeminent cloud computing and big data service provider, holding a leading market share with its state-of-the-art server products
Advertisements
They have also demonstrated considerable R&D prowess in computational infrastructure.
Additionally, Haiguang Information specializes in the development and sales of high-end processors used in servers and workstationsIts product line includes both general-purpose chips and co-processors, indicating a targeted focus on the multi-faceted demands of the computational landscape.
An interesting case study is the capital online brand, which has developed cloud computing nodes across over ten countries, positioning itself to provide computational resources for various applications, such as cloud gaming, AI, and smart manufacturingMoreover, Hongxin Electronics is diligently expanding its computational capabilities in two key areas: partnering with AI chip manufacturers to produce computational hardware, and launching a rental service tailored to the demands of clients needing high-performance models across diverse industries.
Data from mid-2024 indicates a healthy landscape for companies engaged in computational services, with numerous A-share companies reporting substantial revenues from their computational sectorsHongxin Electronics reported over $1.5 billion in revenue from its computational operations, marking it as a leader in this nicheCompanies like Yacorn and Runjian also reported impressive figures exceeding $100 million from their respective computational business segments, signifying a trend among firms to strategically invest in computational infrastructure as a means of long-term growth.
Data centers notably play a pivotal role in the computational capabilities infrastructureReports indicate that 12 firms disclosed their revenues from Internet Data Center (IDC) operations, with Leyou Technology and Huayun New Network each surpassing the $1.5 billion mark in IDC revenueIn terms of proportional revenue, companies like DataPort and CityLand Hong Kong reported IDC revenues accounting for over half of their total incomeLeyou Technology’s combined IDC and Artificial Intelligence Data Center (AIDC) revenues formed an impressive 100% of its operational output, showcasing the evolution of data-centric business strategies in today’s market.
Advertisements