On July 14, 2026, two seemingly unrelated announcements brought to light two main lines of the current AI competition: one focused on "who is considered an AI company," and the other on "who can truly afford AI computing power." The Hong Kong-listed company, Knowledge Atlas (02513.HK), confirmed in an announcement on the Hong Kong Stock Exchange that its name had officially changed from the somewhat obscure Knowledge Atlas Technology Joint Stock Company Limited to the distinctly pointed Z.AI Co., Ltd., while the Chinese name, 北京智谱华章科技股份有限公司, remained unchanged. This move to incorporate "AI" into the company's English name was widely interpreted as a way to strengthen the artificial intelligence label in the capital markets and public perception, riding the wave of the trend toward brand "AI transformation." On the same day, cloud computing service provider Nebius Group announced that it had signed a power sale agreement worth over $1 billion with Reflection AI, aiming for a contract term that stretches to 2029. Under this long-term agreement, the latter would secure large-scale computing resources based on NVIDIA's GB300 AI chips. Public information did not show any direct business intersection between Knowledge Atlas and Nebius or Reflection AI in this agreement, but the timing of both announcements clearly reflected two key battle lines in the industry: on one side, striving to claim the title of "AI company" through name changes and brand reshaping, and on the other side, locking in base resources and cost expectations through billion-dollar, multi-year computing power contracts. Together, these two battle lines formed the basic framework for the subsequent AI arms race.
Z.AI debuts on the Hong Kong Stock Exchange: A name change aimed at AI minds
On July 14, 2026, the name change announcement submitted by Knowledge Atlas to the Hong Kong Stock Exchange appeared to simply compress a lengthy English name—"Knowledge Atlas Technology Joint Stock Company Limited"—into the minimalistic "Z.AI Co., Ltd." However, the announcement deliberately emphasized two unchanged points: the company's Chinese name remains "北京智谱华章科技股份有限公司," and the Hong Kong stock code is still 02513.HK. This means there was no structural adjustment to the listing entity, equity, or regulatory framework; only the English title displayed on trading screens and presentation materials changed, representing a rewrite of external perceptions rather than a restructuring of internal governance.
Simplifying the English name to "Z.AI" leaves almost no room for interpretation: the ultra-minimal symbol Z paired with ".AI" directly stamps the "artificial intelligence company" label onto the company name. The previous name "Knowledge Atlas" resembled a metaphor for a technological path; now, "Z.AI" serves as a clear announcement aimed at the capital markets and the public—Knowledge Atlas hopes to be classified as an "AI-native" entity at the forefront of this arms race. In the context of the past two years, where global technology and internet companies have increasingly integrated "AI" labels into their branding and product lines, this name change is almost a textbook example: it avoids touching the legal entity and securities code, completing a high-recognition brand "AI transformation" at the lowest institutional cost, and seizing a collective imagination portal concerning "who is an AI company" in an information-overloaded market.
$1 billion long-term contract: Nebius sells computing power through 2029
On the same day, Nebius, on the other side of the stage, did not change its name but presented a contract. Nebius Group announced that it had reached a computing power sale agreement worth over $1 billion with Reflection AI, with the contract extending directly to 2029. This is not a temporary rental on a pay-as-you-go basis; rather, it is a bulk pre-sale of computing capacity for the future years: Nebius locks in multi-year incomes and data center utilization rates, while Reflection AI exchanges a long-term contract for predictable computing supply and cost curves, nailing down the most difficult variable in the expiring budget for long-term model training.
The more critical annotation of the agreement is written in the technology stack—Reflection AI will obtain computing resources based on the NVIDIA GB300 AI chips through this long-term contract. Nebius's role is to build large-scale computing infrastructure for AI startups, but the real high point in the industry chain is NVIDIA, which provides the GB300. This long contract nominally occurs between the cloud service provider and the AI startup but essentially quantifies industry's dependence on high-end AI chips as a "$1 billion demand by 2029." As long-term computing power procurement contracts gradually become standard actions for AI companies and cloud service providers to secure resources and price expectations, whoever controls these types of chips holds the decision power over budget sheets in the arms race.
From names to data centers: AI companies vying for positions on two fronts
On the same July 14, 2026, Knowledge Atlas (02513.HK) changed its English name from Knowledge Atlas Technology Joint Stock Company Limited to Z.AI Co., Ltd., attempting to incorporate "AI" directly into its company title; on the other side, Nebius Group confirmed its $1 billion computing power sale agreement with Reflection AI targeting 2029. Two unrelated announcements were awkwardly spliced together by timeline, outlining two opposing battle lines for current AI companies: claiming the title of "AI company" in capital markets and user perceptions on one front while vying for sustainable computing power on the foundational infrastructure front.
For Knowledge Atlas, changing its English name to Z.AI is a standard external narrative action, aiming for investors and users to immediately think "this is an AI company" upon seeing the code 02513.HK, falling into the realm of brand "AI transformation"; for Reflection AI, securing multi-year computing power based on NVIDIA GB300 AI chips through Nebius represents a fundamental internal productivity issue—whether the model can continue to train and whether the reasoning costs can be manageable are tied to this contract. Nebius, as a cloud computing service provider, stands in the middle by packaging high-performance chips like the GB300 as long-term supplies, but the true bargaining power along this line remains with upstream chip suppliers like NVIDIA: when large model development inherently requires long-term and expensive computing investments, those who control the chip production capacity can set the baseline for the next stage of competition outside of brand stories and computing contracts.
Capital-intensive era: The signal from a $1 billion computing power bill
Nebius and Reflection AI have locked in over $1 billion based on NVIDIA GB300 chips for computing agreements before 2029, starkly writing the entry threshold for cutting-edge AI projects into a multi-year massive bill. For startups like Reflection AI, this is not just "buying servers," but rather pre-mortgaging their financing capacity and product monetization rhythms for several years to the computing supply party: a long-term contract provides certainty regarding resources and price expectations while also documenting high-intensity computing investments as nearly irreversible fixed costs. The industry-wide consensus that "without sustained large computing power, there is no competitiveness" crystallizes into cash flow pressure and changes in life-and-death rhythms on this $1 billion contract—as soon as a technical route is chosen, one must constantly present larger business stories to cover this computing expense.
From Nebius's perspective, this contract is also a bet. Long-term computing power procurement contracts are becoming the mainstream model for securing resources between AI startups and cloud service providers, allowing Nebius to lock in multi-year income and cluster utilization rates while increasing its confidence to place larger orders with upstream chip suppliers like NVIDIA, thus diluting its initial investments; however, it also ties itself to the GB300 production capacity and pricing cycle, as well as the operating conditions of a single large customer. Risk is redistributed between both parties: startups bear the losses of "what if growth does not meet expectations," while cloud service providers undertake the asset risks of "what if computing cycles reverse." Looking back at the same day Knowledge Atlas' rebranding as Z.AI (02513.HK) was announced, the capital market sees two concurrently unfolding battle lines: one through brand "AI transformation" seizing labels, and the other through significant, long-term computing contracts proving their real position at the forefront of the competition. As a listed company on the Hong Kong Stock Exchange, Knowledge Atlas's major capital arrangements must be disclosed in financial reports and announcements, and currently, there is only information about the name change; this means investors, while assessing Knowledge Atlas and other Hong Kong-listed AI targets, must listen to how they narratively define their identity as "AI companies" while also keeping an eye on whether similar long-term computing commitments like those of Nebius and Reflection AI appear in the financial statements, as this is becoming a key variable that distinguishes storytelling companies from those truly entering the capital-intensive race.
The next scene: More Z.AI and more $1 billion orders?
From Knowledge Atlas's name change to Z.AI to Nebius-Reflection AI's $1 billion computing power agreement, July 14, 2026, seems to have handed the industry a screenplay: on one hand, companies writing "AI" across their signs, and on the other, inscribing long-term debts for computing power on their balance sheets. It is foreseeable that more technology and internet companies will follow the earlier path, strengthening their AI identities through rebranding, brand reshaping, and business restructuring, in order to gain higher premiums in capital market narratives; meanwhile, the latter path will likely shift from "news" to "norm"—more Nebius-like cloud service providers will sign multi-year, billion-dollar computing contracts with more Reflection AI-like model companies, and competition over high-end chips like the NVIDIA GB300 will thrust pricing power and supply stability to the forefront. For investors and industry observers, three lines of clues have emerged that are worth closely monitoring in the next phase: who is aligning their company names and product lines toward ".AI / AI," who is securing multi-year computing procurement commitments in announcements and financial reports, and to what extent these contracts are subject to the tightening supply patterns of chip providers like NVIDIA, as these three signals will define the segmentation of player tiers in the future public information.
Join our community to discuss and grow stronger together!
Exclusive Hyperliquid benefits for AiCoin: https://app.hyperliquid.xyz/join/AICOIN88
Exclusive Aster benefits for AiCoin: https://www.asterdex.com/zh-CN/referral/9C50e2
On-chain Telegram community: https://t.me/AiCoinWhaleData
On-chain community: https://www.aicoin.com/link/chat?cid=N6OVMor5g
AiCoin on-chain Twitter: https://x.com/aicoinwhaledata
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
