This week.$NVIDIA (NVDA.US)$CEO Jensen Huang delivered a significant keynote speech at CES, but investors in the stock market were not impressed. On the first trading day after Huang's speech, Nvidia's stock price fell by over 6%, marking the worst single-day performance in four months. Investors felt that Huang's speech was just "fluff" with no substantial content, and believed that there were no updates on Nvidia's Datacenter Business, which is currently of great concern to the market, during this CES.

However, Wall Street thinks the market may be overly worried. On January 7, local time, Analysts attended NVIDIA's Q&A session during CES and subsequently released a report collectively expressing their optimism about NVIDIA's future.

NVIDIA CEO Jensen Huang and CFO Colette Kress hosted this analyst Q&A session and answered various questions from the analysts. The key points are summarized as follows:

NVIDIA launched the AI supercomputer Project Digits, an AI workstation aimed at developers (similar to a mini DGX), which uses the latest GB10 Grace Blackwell super chip.

NVIDIA emphasized that accelerated computing is the direction of future development, stating that the platform advantage of its commercial GPU solutions is the main competitive edge over ASICs.

NVIDIA is still in the early stages of growth. Machine learning is the direction for future development, which will lead to accelerated computing replacing general computing. Therefore, approximately 1 trillion dollars of general computing installation bases will need upgrades in the coming years.

Management expects that in the first quarter of 2025, the revenue from the Blackwell series chips will exceed the previous expectation of "several billion dollars," and at the same time, the combined revenue from Hopper and Blackwell series products will also see growth in that quarter.

Supply chain delays are more caused by an imbalance between demand and supply and system complexity, rather than just supply issues. Nevertheless, the research and production of the Blackwell series systems are progressing fully as planned.

UBS Group believes that the Digits product provides another avenue for NVIDIA to further strengthen its moat in the AI sector. The institution expects revenue to reach 133.2 billion dollars in fiscal 2025 and remains optimistic about NVIDIA's performance in the fourth quarter of fiscal 2025 and the first quarter of fiscal 2026, predicting that Blackwell's shipments for the quarter will be around 14 billion dollars.

Citi believes that NVIDIA's growth is sustainable, driven by four trends: the shift of the traditional computing market towards accelerated computing, AI becoming a new computing layer, demand for AI factories, and AI empowering business productivity.

JPMorgan expects NVIDIA's Datacenter revenue to continue to grow until 2025, benefiting from increased capacity of Blackwell products and market demand. Although Blackwell production is proceeding as planned, the market supply and demand relationship may still be tight after 2025.

UBS Group is bullish on NVIDIA: market concerns are overstated, AI moat solidified, with revenue growth to 133.2 billion dollars in fiscal 2025.

NVIDIA reiterated that it will confirm revenue when its Blackwell compute trays are delivered to original equipment manufacturer (OEM) and original design manufacturer (ODM) partners.

Although some ODM partners cannot absorb large inventories due to balance sheet constraints, UBS Group believes that large hyperscale customers may provide transitional financing to cope with delays in NVL72 and other rack SKUs.

"Therefore, market concerns about a 'gap' or 'slight overperformance' in performance for the fourth quarter of fiscal 2025 (January) and the first quarter of fiscal 2026 (April) are overstated."

UBS Group remains very bullish on NVIDIA's performance in the fourth quarter of fiscal year 2025, expecting Blackwell shipments of approximately $14 billion this quarter, and holds an optimistic attitude towards the guidance for the first quarter of fiscal year 2026.

NVIDIA also launched DIGITS during CES, which is its first desktop product, utilizing the latest GB10 Grace Blackwell superchip, integrating Arm architecture-based CPUs and GPUs, and capable of delivering up to 1 petaflop (one quadrillion floating-point operations per second) of AI computing performance. UBS Group believes this release was somewhat unexpected. Although it is not expected to significantly boost revenue, its strategic importance lies in further solidifying NVIDIA's moat in the AI field.

Regarding the balance between commercial GPUs and custom application-specific integrated circuits (ASICs), NVIDIA first points out that this clearly indicates accelerated computing is the direction for future development. Specifically, NVIDIA emphasizes the platform advantages its commercial GPU solutions offer, which is its main competitive edge over ASICs. The company further notes that AI will become part of the national infrastructure, and accelerated computing is the future trend.

NVIDIA emphasizes that current AI development is still in a very early stage, and accelerated computing will gradually replace general computing. In the coming years, about $1 trillion in global general computing installations will require upgrades.

UBS Group maintains NVIDIA's Target Price at $185, and the current after-hours price of the company is $138.54, which has about a 30% appreciation potential compared to the Target Price. The revenue for fiscal year 2025 is expected to be $133.243 billion, with EBIT at $88.792 billion, net income at $75.814 billion, and diluted EPS at $3.05, significantly higher than $1.30 for fiscal year 2024.

Citigroup maintains a 'Buy' rating on NVIDIA, focusing on short-term Datacenter growth and long-term AI-driven potential.

Citigroup states it retains a 'Buy' rating on NVIDIA, with a Target Price of $175, indicating a potential increase of 24.9% over the current stock price, and expects a total ROI of 24.9%. The main reasons include the optimistic outlook for short-term datacenter revenue growth and the sustainability of long-term growth.

Management expects the Blackwell series chips to exceed the previous revenue guidance of "several billion dollars" in the first quarter of 2025. Additionally, the combined revenue of the "Hopper" and "Blackwell" product lines will also see growth in that quarter.

Citigroup stated that NVIDIA believes its growth is sustainable and outlined four major trends driving this growth:

• The tail effect in the traditional computing market: NVIDIA believes that as more workloads shift to accelerated computing, the company will continue to benefit from the $2 trillion traditional computing market.

• AI as a new layer of computing: AI not only enhances existing tools but also becomes a new computing infrastructure.

• Demand for AI factories: The presence of AI services will require significant capital expenditure to build "AI factories", which will be bullish for NVIDIA.

• AI empowering enterprise productivity: AI agents (such as digital employees) will drive business efficiency.

NVIDIA indicated that progress has been made in addressing supply chain delays, with improved relationships with major suppliers and collaboration with new suppliers. The company emphasized that the delays are more a result of imbalance between demand and supply and systemic complexity rather than solely supply issues. Nevertheless, the "Blackwell" series systems are still progressing on schedule.

NVIDIA's management pointed out that discussions about GPUs and ASICs indicate that accelerated computing has become a new development direction. They believe their GPU solutions outperform ASICs in scale and aim to create a general computing platform to address the next wave of AI applications. Furthermore, national-level AI infrastructure construction will also require NVIDIA's solutions, which ASICs cannot meet.

Autonomous driving requires a significant amount of data processing and model training. NVIDIA believes that all electric vehicle companies need to be equipped with AV platforms, and its solutions are at the core of this field. Edge AI and Siasun Robot&Automation technologies are two other important AI application areas.

NVIDIA sees opportunities in training, which presents a huge datacenter opportunity. Companies in the robotics field face the challenge of a lack of data, so NVIDIA provides a data generation platform within NVIDIA Omniverse. The company also hopes to drive industry development through simulation, with management believing this is its most advantageous area.

Citigroup has set a target price of $175 for NVIDIA, based on an estimated EPS for fiscal year 2025, giving a PE of approximately 35 times. This multiple is close to its one-year historical average but lower than the three-year and five-year averages.

JPMorgan: AI and accelerated computing drive NVIDIA's trillion-dollar market opportunities, target price raised to $170.

JPMorgan's research team expects spending on AI and accelerated computing to remain strong over the next decade and beyond. NVIDIA's revenue growth in the datacenter field is expected to continue until 2025, largely due to capacity increases in the next-generation Blackwell products and strong market demand.

In the long run, NVIDIA will achieve significant revenue growth by capturing a larger share of the $1 trillion datacenter infrastructure market. This trend is driven by the transformation into accelerated computing and the rising demand for AI solutions.

JPMorgan stated that the production of Blackwell-related products is on schedule and is expected to support revenue growth in 2025. However, even after entering 2025, the market's supply-demand relationship may remain tense, particularly in areas such as advanced packaging technology, advanced memory, and networking equipment.

To address supply chain challenges, the NVIDIA team is closely collaborating with supply chain partners to gradually alleviate bottleneck issues. For example, the GB200 and Ultra product lines are expected to see capacity increases in late 2025.

The complexity of training models, new scaling laws, and the demand for test time calculations are expected to continue driving the long-term demand for GPU computing power. JPMorgan noted that as the share of inference and long-term thinking in AI inference applications increases, the calculation of test time becomes crucial.

NVIDIA's solutions have significant advantages in system integration, software stack, ecosystem, and ease of use. This comprehensive system solution makes it outperform custom ASIC solutions in performance. JPMorgan expects that enterprises, vertical markets, and sovereign clients will continue to favor NVIDIA's technology.

Agent AI is considered a key driver of growth in demand from NVIDIA's enterprise clients. To accelerate the application of Agent AI, NVIDIA has launched Blueprints solutions tailored for specific business scenarios, such as call centers, fraud detection, and risk management. The report estimates that this field will positively contribute to revenue growth for enterprises and vertical clients.

At the 2025 CES, NVIDIA debuted the next-generation Blackwell GeForce RTX 50 series, which offers approximately double the performance of the previous generation. Additionally, NVIDIA released a small supercomputer focused on AI projects called Project Digits. JPMorgan stated that these initiatives demonstrate NVIDIA's strong position in the high-end gaming and AI PC markets, providing a good opportunity for its expansion from the high-end gaming market to the consumer market.

JPMorgan rated NVIDIA as "Shareholding" with a Target Price of $170.

Want to enter the market but find timing difficult."Monthly Payment Zone"The feature helps you to invest regularly and steadily, capturing investment opportunities!

Editor/ping