On April 22, 2025, the publicly offered "Champion Fund" - Everwin Technology Intelligent Selection disclosed the first quarter report for 2026. In terms of performance, as of the end of the reporting period, the net asset value of the A Fund of Everwin Technology Intelligent Selection mixed fund was 3.7494 yuan, with a net asset value growth rate of -0.80% during the reporting period, while the benchmark return rate for the same period was -4.51%; as of the end of the reporting period, the net asset value of the C Fund of Everwin Technology Intelligent Selection mixed fund was 3.7169 yuan, with a net asset value growth rate of -0.94% during the reporting period, while the benchmark return rate for the same period was -4.51%. As of the end of the first quarter of this year, the top ten heavy-weighted stocks selected by Everwin Technology Intelligent Selection were: Inter China Creation (300308.SZ), New Yisheng (300502.SZ), Hudi Electric (002463.SZ), Shennan Circuit (002916.SZ), Dongshan Precision (002384.SZ), Tengjing Technology (688195.SH), Jiangfeng Electronics (300666.SZ), Boqian New Material (605376.SH), China Giant Stone (600176.SH), and Dingtai High-Tech (301377.SZ). In terms of portfolio adjustment, Everwin Technology Intelligent Selection added Tengjing Technology, Jiangfeng Electronics, Boqian New Material, and China Giant Stone in the first quarter, and significantly increased its position in Dingtai High-Tech by 123.79%, making it one of the top ten heavily-weighted stocks. These five assets have all shown an increase in their value by over 70% this year, with Tengjing Technology rising by over 130% and becoming the "dark horse" in the portfolio. At the same time, the originally most heavily-weighted stocks such as Shengyi Technology, Tianfu Communication, Jingwang Electronics, Industrial Fulian, and Cambridge Technology exited the top ten. Among the remaining heavily-weighted stocks, the fund continued to increase its positions in Inter China Creation, New Yisheng, Hudi Electric, and Shennan Circuit, and only slightly reduced its position in Dongshan Precision. Fund manager Ren Jie analyzed in the first quarter report that the Shanghai and Shenzhen 300 Index fell by 3.89%, the CSI 1000 Index rose by 0.32%, and the Kechuang 50 Index fell by 6.54% in the first quarter of 2026. During the reporting period, the fund continued to focus on investment in the global cloud computing industry, tracking the forefront models and commercialization progress of the global AI industry, and investing in response to key areas of supply. In the first quarter of 2026, Token entered the explosive period of the third phase. Ren Jie emphasized that with the maturity of AI Agent technology represented by Anthropic, Agents are no longer just simple prompt word calls, but truly autonomous intelligent entities that can plan, decompose, and execute multi-step tasks, and maintain goal consistency in long-term work. Outside of Agents, there is still significant potential for improvement in future models in areas such as multimodality and continuous learning. In terms of training, Ren Jie recently observed that the new generation of models is being trained on larger parameter scales, and has already shown breakthrough improvement capabilities, and new models from companies like Meta also verify the potential for further scaling. The explosion of Token has already been manifested in performance and valuation. Two leading model startups have approached a 30 billion ARR, and some domestic companies have also achieved significant revenue growth. The financing and valuation of various assets in the primary and secondary markets reflect the continued optimism of global investors on the long-term development of models. Ren Jie pointed out that the architecture of computing power is adapting to the changing demands of models and applications. From the perspective of customer demand, the training of reinforcement learning, Agent tasks, MAS/long-context and other different architecture requirements are all placing higher demands on inference speed.The providers of computing power are beginning to optimize the Decode stage further after PD separation. During this process, the importance of communication and storage in computing systems is rapidly rising, and in order to cope with latency, bandwidth, and energy efficiency constraints, new technologies such as LPU, CPO/NPO, OCS, and orthogonal backplanes are being rapidly introduced into the new generation of computing clusters. As new technologies move from solution design to large-scale deployment, companies that have early deep involvement in top-tier manufacturer R&D and supply chain collaboration are expected to gain more significant industry dividends as penetration rates increase. The demand for computing power is expanding into more industries. As AI continues to expand in scale, the overall demand for more segmented industries is significantly driven, with high-end demands in certain areas occupying more production capacity, and some areas entering a significant price increase stage under supply-demand mismatch, the market situation is also becoming more generalized. Overall, global AI models are still in a phase of continuous improvement in capabilities and continuous expansion of application scenarios, with strong sustainability and certainty in industry development. This fund will continue to focus on cloud computing, and will focus on changes in the AI industry to capture investment opportunities brought about by AI construction and development.