On February 10, 2025, a powerful force seemed to have ignited the A-share market as DeepSeek-related stocks surged once againAmidst the clamor and excitement, the market sentiment reached a peak, driven by the groundbreaking announcement that the three major telecommunications companies had fully integrated with the DeepSeek open-source large modelThis integration has transformed the landscape of artificial intelligence (AI) investments, and the ripple effects are immediate and significant for investors and companies alike.

The share prices of the three telecommunications giants soared as if injected with a potent boosterChina Telecom’s stocks skyrocketed, reaching their limit and setting a historical record, underscoring the market's optimistic outlook on this collaboration; China Unicom did not lag behind, also surging more than 7%, reflecting significant market vitality; China Mobile garnered attention, with gains exceeding 4%. Besides these major players, many associated stocks also rose sharplyMegvii Technology hit a remarkable six consecutive trading limits, becoming the focal point of the market's interest; Qingyun Technology and Hangzhou Steel Group each recorded four consecutive limits, showcasing impressive upward momentum; YK Technology and Wan Shun New Materials saw their stock prices soar by over 10%, further heating up the market

Advertisements

By the time this report was drafted, while the gains had relaxed somewhat, the fervor of this movement remained striking.

Following the Spring Festival, the first trading week saw DeepSeek stocks maintaining their lead in the A-share marketFrom February 5 to 10, specific segments such as AI data, computing power leasing, and edge-side AI rose sharplyFinancial inflows reflected this trend; an AI ETF saw a net inflow of nearly 20 million yuan over the past week, indicating strong investor interest and demand in the AI sector, further driving the market's momentum.

The surge in market activity correlates closely with the commercialization advancements of the DeepSeek modelOn February 6, China Mobile, China Unicom, and China Telecom jointly announced the successful integration of DeepSeek-R1 and V3 models on their cloud computing platformsThis strategic initiative is pivotal, as it enhances the practical applications of the DeepSeek model, leveraging the telecommunications companies' formidable cloud computing capabilitiesLooking ahead, the partners aim to collaborate on deploying this model across various sectors such as smart cities and industrial quality inspections, injecting fresh energy into their developmental trajectoriesPreviously, companies like Baidu Cloud, Tencent Cloud, and Huawei Cloud also introduced cost-effective solutions based on the DeepSeek model, employing time-limited free strategies, which act as a catalyst for engaging market needs and encouraging wider adoption of DeepSeek's offerings.

The explosive growth of DeepSeek is not merely fortuitous; it is a result of significant technological innovationsReports indicate that the DeepSeek-V3 model has achieved remarkable advancements through intricate algorithmic optimizations, cutting training costs down to 10% of the industry average, enabling more businesses to participate in AI developments affordably

Advertisements

Moreover, its inference efficiency has improved by more than three times, achieving performances in critical tasks such as code generation and multimodal understanding that rival GPT-4. This technological leap is poised to reshape the competitive landscape in the global AI arena.

Experts analyze that the “low-cost + high-performance” combination of DeepSeek is groundbreakingThis development allows small and medium-sized enterprises to bypass longstanding constraints regarding computing power, enabling them to engage actively in the AI application development waveForecasts suggest that by 2025, the pace of commercialization for China's AI large models could accelerate by 40%, significantly bolstering the country's AI industryFurthermore, DeepSeek's influence is extending into international markets; global giants like Amazon and NVIDIA have announced their adoption of the DeepSeek model, indicating its widespread recognition and application on a global scaleMeanwhile, reports indicate that Meta is urgently initiating “anti-DeepSeek” technological initiatives internally, reflecting the model's impact and the competitive pressure it exerts on industry behemoths.

However, as enthusiasm in the market continues to rise, subtle risk signals have begun to surfaceIn the capital market, several companies have issued statements clarifying that they do not hold equity in DeepSeek's parent company nor have they provided training data, emphasizing that business cooperation has many uncertaintiesThis situation suggests that some stocks tied to DeepSeek may be experiencing excessive speculation, leading to price increases without tangible business backingMoreover, the rise in phishing websites impersonating DeepSeek has surged, with scammers advertising “income of one million a month” to sell related courses and exploit investors

Advertisements

Advertisements

Advertisements