After ChatGPT kicked off the AI 2.0 era, everyone was catching up with big models.
If data is the “fuel” of an artificial intelligence model, then computing power is its “engine.” As a result, the cloud services market has also entered a new 2.0 phase. However, different cloud computing vendors have chosen different development paths. As the largest public cloud vendor, Alibaba Cloud developed its own Tongyi series of big models; Baidu Cloud used Wenxin's words to try to fight a beautiful turnaround battle; HUAWEI CLOUD was also one of the earliest cloud service providers in China to lay out a large-scale model, and already released the Pangu series of large-scale pre-training models as early as 2021.
But not all cloud computing vendors are using an integrated model to compete.
For example, Jinshan Cloud adheres to the position of a “neutral cloud service provider” and focuses on digging deeper into the value of cloud computing, helping various industries build large-scale models, and jointly layout the field of artificial intelligence with ecological enterprises. The company's recently released financial report shows that Jinshanyun is gradually embarking on its own unique development path. As of 2023, Jinshan Cloud's artificial intelligence revenue increased sharply by 82% month-on-month, and AI-related businesses have become the company's new growth engine.
The management of Jinshan Cloud pointed out during the earnings conference call that they have noticed a trend of increasing demand for artificial intelligence computing power. As the training parameters of old customer models improve, their demand for computing power is also increasing. In addition to the expansion needs of existing customers, new customers also want to incorporate the capabilities of large models into their existing business. Management said, “As artificial intelligence technology continues to penetrate the hearts of the people, and people's awareness and acceptance of it continue to increase, there will be huge room for new customers to grow in the future.”
Director Yangji sticks to the position of “big model supporter”
According to the performance report, under the premise that the company actively reduced the scale of the CDN business and screened industry cloud projects more strictly, Q4 Jinshan Cloud still achieved revenue of 1,722 billion yuan, an increase of 6.0% over the previous month. During the reporting period, the company's adjusted gross margin reached 15.2%, achieving a record high for the sixth consecutive quarter; gross profit of 253 million yuan, a year-on-year increase of 56.3%; and the normalized adjusted EBIDTA ratio was -1.6%, a sharp decrease of 8.6 percentage points from the same period last year. These all show that the company's profitability is constantly improving, towardsEstablished goalsMove forward steadily.
Quantitative change will bring about qualitative change, and this turning point of “qualitative change” may be imminent.
Recently, a number of major banks, including J.P. Morgan, Citibank, and CITIC Lyon, raised their company ratings. Everbright Securities Research Institute predicts that with the continuous optimization of the public cloud revenue structure, the improvement of the quality of industry cloud projects, and the continuous advancement of cost reduction and efficiency, the company's profitability will continue to improve. J.P. Morgan Chase further predicted in the research report that Jinshan Cloud will achieve profit before interest, tax, depreciation and amortization (EBITDA) break-even in the first quarter of '24, and achieve full year EBITDA balance for the first time in 2024 (while the consensus expectation is an EBITDA loss).
Outstanding financial results are inseparable from Jinshanyun's tight grip over the past yearAIThe opportunity to do a good job leading the way”big modelhelpers”.
On the one hand, Jinshan Cloud remains steadfast in its “neutral” position, and has not joined the Hundred Regiment Wars with the trend to avoid forming competitive relationships with customers — this is very rare in the cloud computing industry. Instead, the company is committed to providing high-quality cloud services to customers in the fields it focuses on, becoming a loyal partner in the field of artificial intelligence for ecological enterprise layout, and acting as an “upstream and downstream linker”.
On the other hand, Jinshan Cloud has been deeply involved in the cloud computing industry for more than ten years and has established a mature technical system, business model, and solution experience. During the reporting period, the company consolidated and expanded its development advantages from various aspects around AIGC, and contributed to the next stage of development. The company set up an artificial intelligence R&D center to build technical capabilities for the needs and characteristics of large models, provide stable and efficient support for training, deployment and reasoning of large models, continuously upgrade its core storage, database, network and other products, and continuously improve MaaS mutual trust zone solutions to provide customers with better artificial intelligence infrastructure and platform services. At the same time, Jinshan Cloud is also actively cooperating with leading enterprises in various fields to promote the development and application of artificial intelligence technology to achieve win-win cooperation.
Helping industrial digitalization stimulate new momentum for growth
Jinshan Cloud became a “supporter” rather than a “developer” of high-quality large-scale models, and the results are obvious — the company will join hands with a broad customer base and benefit from the wave of artificial intelligence sweeping all walks of life.
For example, Jinshan Cloud is fully benefiting from the rapid advance of ecological partners' artificial intelligence processes. During the reporting period, Jinshan Cloud's ecological strategic cooperation with Millet & Jinshan was further deepened. In 2023, Jinshan Office released WPSAI, an intelligent office application based on the big language model, which anchors the development of the three strategic directions of AIGC (content creation), Copilot (smart assistant), and Insight (knowledge insight), bringing the results of AI pioneering implementation in the domestic office software field to users. Among them, WPS AI began public testing in Q4 2023 to prepare for large-scale commercialization in the future. On the Xiaomi side, the Xiaomi SU7 will be officially launched on March 28. Judging from publicly available information, Xiaomi cars are equipped with highlights such as a surging OS with cross-end connectivity and the support of large AI models, so there is a demand for cloud use for training and inference. Jinshan Cloud has grasped the core needs of ecological partners. In Q4 2023, revenue contributions from eco-partner companies reached 16%, an increase of 5 pcts over the previous year. Based on the progress of Eco Partner's business, it is easy to guess that this data is expected to perform better in 2024.
In addition, Jinshan Cloud uses high-quality government and enterprise customer resources to tap higher customer value. Q4. On the basis of maintaining stable cooperation with existing major customers, Colette signed 4 new customers. Furthermore, according to media reports, Jinshan Cloud has established a series of industry cloud benchmarks such as Beijing Government Cloud, Haidian Education Cloud, Chibi Government Cloud, etc., and provides services to Shenzhen State-owned Cloud, Rizhao State-owned Cloud, and Shanghai Digital Production Cloud, becoming the digital intelligence foundation for digital government. Jinshan Cloud is gradually becoming a new type of productivity that helps governments and enterprises break away from traditional growth paths and meet the requirements of high-quality development.
According to the Everbright Securities Research Institute, big model training and inference for Internet customers, intelligent driving training for car companies, and data migration to the cloud are all expected to drive AI computing power demand and predict that the AI business is expected to continue to contribute to revenue growth in the future.
epilogue
It is easy to see that a series of financial data and operational data prove that Jinshan Cloud is beginning to realize the potential of artificial intelligence and big models.
Cloud computing is very deterministic as the underlying infrastructure in a high-boom large-scale model circuit. The big model is like a “gold digger,” and the huge potential is also full of risks and uncertainties - “gold diggers” may not only get nothing after trying; even if they find “gold,” if there are too many “gold diggers” in the gold mine, the competition will become very intense. However, in the midst of the gold rush boom, the business of selling shovels has a strong degree of certainty.
Therefore, if you look for a “big model helper” with strength and potential, you will have a greater chance of winning. This may be the reason why Jinshanyun is favored by the capital market.