NVIDIA Corporation (NVDA.US) GTC Berlin signals 2027 growth! Morgan Stanley positive on supply chain expansion and AI factory efficiency gains, reiterates top pick.

Morgan Stanley released a research report after attending NVIDIA Corporation's (NVDA.US) GTC Berlin conference this week, stating that although AI infrastructure construction is facing multiple challenges such as chip supply, data center power, and financing, NVIDIA Corporation is further consolidating its competitive advantage by expanding supply chain capacity, optimizing AI data center deployment efficiency, and advancing the next-generation Vera Rubin platform. The bank believes that NVIDIA Corporation is expected to maintain strong growth in 2027, reiterates its position as the top pick in the semiconductor industry, and maintains an "Overweight" rating and a $300 price target. Based on the closing price of $221.97 on October 21, Morgan Stanley's price target implies approximately 35% potential upside for NVIDIA Corporation's stock price. Morgan Stanley analyst Joseph Moore noted that although the GTC Berlin conference did not disclose many new details about the next-generation Rubin Ultra products, NVIDIA Corporation's progress in supply chain expansion, AI data center deployment, and software ecosystem further strengthened the bank's confidence in the company's 2027 growth prospects. Morgan Stanley expects NVIDIA Corporation's revenue to grow by approximately 88% and 70% in 2026 and 2027, respectively. The bank noted that NVIDIA Corporation management had previously described demand growth of approximately 100%, significantly higher than the approximately 70% revenue growth forecast for 2027, reflecting that supply capacity remains an important factor constraining further revenue growth. Therefore, how to narrow the gap between market demand and actual delivery capacity has become key to NVIDIA Corporation's future growth. The bank believes that NVIDIA Corporation is gradually alleviating supply chain bottlenecks by expanding chip substrate, advanced packaging, and testing capacity. Among these, Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR's 3nm wafer capacity remains the main supply constraint in the bank's view, but the related pressure is expected to gradually ease over time. According to the research report, NVIDIA Corporation is expanding cooperation with substrate suppliers such as IBIDEN, SEMCO, and Unimicron, and related capacity is expected to expand to 2.5 times the current level over the next three years. In the advanced packaging field, Amkor plans to increase CoWoS-R wafer packaging capacity by more than 300% by the end of 2028 to support the growth of CPU and other processor products. On the testing side, NVIDIA Corporation has committed $1.1 billion to reserve capacity at KYEC's new facilities and equipment in the United States, and the related facilities are expected to enter mass production in April 2027. NVIDIA Corporation also expects its overall semiconductor testing capacity to double in 2027 compared with 2026. In addition, in the co-packaged optics (CPO) field, SPIL, under ASE, plans to expand related capacity to 15 times the current level by mid-2027 and expects to further double it on that basis by the end of 2028. Morgan Stanley noted that these supply chain investments not only help improve NVIDIA Corporation's product delivery capacity but may also enable the company to further expand its competitive advantage against the backdrop of continued rapid growth in AI infrastructure demand. In addition to supply chain expansion, the DSX POD reference design launched by NVIDIA Corporation at GTC Berlin has also become a key technology development for Morgan Stanley. DSX POD is a modular AI data center design solution for Vera Rubin systems. The related equipment can be integrated and validated before delivery, enabling customers to put it into operation faster after receiving it. Morgan Stanley noted that traditional AI data centers often require months from equipment delivery to officially generating computing revenue, while with DSX POD, this process is expected to be shortened to days. For new cloud service providers that rely on large amounts of debt financing to build AI computing infrastructure, shortening the equipment deployment cycle is of great significance. The bank believes that AI servers and GPUs typically cannot generate revenue before officially being put into operation, yet they already tie up substantial capital. If NVIDIA Corporation can significantly shorten this stage, customers can obtain cash flow earlier, thereby improving working capital turnover and alleviating financing pressure. Morgan Stanley emphasized that the market usually focuses more on GPU performance, computing scale, and product launches, but the time required for AI infrastructure to go from delivery to revenue generation may also become an important factor affecting customers' return on investment. NVIDIA Corporation also introduced the DSX MaxLPS solution. According to information disclosed at the conference, this solution can reduce power consumption by approximately 30% without reducing the token output of AI models. Morgan Stanley believes that against the backdrop of global AI data centers facing power supply and construction land constraints, improving the computing power generated per unit of electricity will become an important way for NVIDIA Corporation to differentiate itself from competitors. NVIDIA Corporation expects that more than 10 customers in Europe will adopt the DSX platform to build Vera Rubin AI factories. In addition to hardware and infrastructure, NVIDIA Corporation is also accelerating the expansion of AI software and other application areas. In the agentic AI field, the company launched the OpenShell reference architecture. This is an open-source framework based on Linux that can isolate different AI agents in independent secure environments and limit their system access permissions. At the same time, Nvidia Sentry running on the BlueField-4 platform can monitor AI agent activity from outside the system, further enhancing security. Morgan Stanley noted that these products help NVIDIA Corporation expand its business scope in the AI agent field from GPUs to Vera CPUs and BlueField network processors, increasing the variety of NVIDIA Corporation products used in each customer deployment. In the communications field, NVIDIA Corporation collaborated with Nokia Oyj Sponsored ADR to complete the first commercial phone call based on CUDA technology on T-Mobile's network. NVIDIA Corporation CEO Jensen Huang also proposed that in the future, approximately 100,000 telecom base stations and data center sites, totaling about 100 gigawatts of infrastructure, are expected to gradually transform into distributed AI computing networks. In autonomous driving, NVIDIA Corporation announced that four additional automotive brands will be compatible with the Hyperion autonomous driving platform, further promoting it as an industry-standard architecture. In the humanoid Siasun Robot&Automation field, NVIDIA Corporation partnered with Agile Robots to launch the Agile ONE SN reference Siasun Robot&Automation, which is expected to begin delivering pre-assembled products equipped with computing and software systems in the first quarter of 2027. Morgan Stanley believes that these developments show that NVIDIA Corporation is expanding its competitive advantage from traditional GPUs to multiple fields such as AI infrastructure, software, communications, autonomous driving, and Siasun Robot&Automation, creating more opportunities for long-term growth. On valuation, Morgan Stanley maintains NVIDIA Corporation's base-case price target of $300, corresponding to approximately 20 times its 2027 calendar-year EPS forecast of $15.01. The bank believes that although NVIDIA Corporation already has a high market share and gross margin, room for further short-term valuation multiple expansion may be limited, while strong AI demand and continued progress on the Vera Rubin platform are expected to drive further upward revisions to earnings forecasts. In the bull case, Morgan Stanley assigns NVIDIA Corporation a valuation of $350, mainly assuming that data center revenue further exceeds expectations as supply improves, while CPUs, AI software, and other new businesses accelerate growth. In the bear case, the bank assigns NVIDIA Corporation a valuation of $160, with main risks including a significant slowdown in AI data center demand growth, competitors launching more competitive products, and large cloud computing customers accelerating adoption of in-house chips. Overall, Morgan Stanley believes that NVIDIA Corporation's core strengths demonstrated at GTC Berlin lie not only in the performance improvements of next-generation GPU products but also in the complete system it is building covering chip supply, advanced packaging, data center deployment, and software ecosystem. As the Vera Rubin product cycle gradually begins and supply chain capacity and AI data center deployment efficiency continue to improve, the bank believes NVIDIA Corporation is expected to further narrow the gap between demand and supply and sustain strong growth momentum in 2027.