On December 28, 2021, we released a publication entitled Nvidia’s Trillion-Dollar AI Opportunity which delved into the reasons behind Nvidia’s (NASDAQ:NVDA) dominance in the AI industry and its potential for growth. Our report was largely theoretical and focused on analyzing the economic, social, and technical factors contributing to the centralization of AI-based business models. The AI market has experienced remarkable growth since the publication of our report, surpassing even our most optimistic projections. In our view, the introduction of ChatGPT to the world has redefined our understanding of AI from theoretical to tangible.
As the dynamic AI industry continues its rapid ascent, Nvidia remains poised to be a major beneficiary of the advancements, particularly with the introduction of ChatGPT. Despite receiving significant attention, the full extent of the potential impact of ChatGPT on NVDA may be underestimated by investors. This article will explore the impacts of ChatGPT on the developments and effects of the AI arms race as it relates to NVDA, as well as providing an analysis of the company’s financials and valuation to give readers a comprehensive understanding of the potential risks and rewards of investing in the stock.
The introduction of ChatGPT has catalyzed an arms race among the largest technology companies. According to media reports, Microsoft, Alphabet and Meta are all gearing up to spend more. These three companies are expected to spend an incredible $90 billion in capital in their current fiscal year.
Over one million users registered for ChatGPT in the first five days after its launch on November 30, 0222, as reported by Greg Brockman, President of OpenAI. The program’s capabilities have sparked a lot of discussion on social media, with users expressing amazement, amusement, and concern.
As reported by NYTimes, ChatGPT’s release prompted Alphabet’s (GOOG) (GOOGL) management to declare a “code red,” indicating a major threat to the company. Some fear that this could be a sign of an impending disruptive technological change that could significantly impact the company’s business. Google’s search engine has been the primary way for people to access the internet for over two decades, but new chatbot technology that could potentially replace traditional search engines could pose a significant threat to Google’s main search business. One executive at Google has described the situation to the NYTimes as a “make or break” for the company’s future.
Microsoft CEO Satya Nadella stated at the World Economic Forum this week that “Every product of Microsoft will have some of the same AI capabilities to completely transform the product,” and tools like ChatGPT are necessary to boost productivity. Microsoft has also obtained an exclusive license to the underlying technology behind GPT-3 in 2020 and has a strong partnership with OpenAI. An update to GPT-3 called GPT-4 is expected to be released in early 2023 and be significantly more powerful than the current version, according to media reports.
The introduction of ChatGPT has also awakened Meta (META) from its slumber as employees have recently shared internal memos urging the company to speed up its AI approval process to take advantage of the latest technology. Moreover, it has been reported that Meta is increasing employing AI to fend up competitors like TikTok.
As the AI arms race heats up, the only company almost guaranteed to be a winner is NVIDIA. We can say this with confidence given NVIDIA’s market share in AI processing is 80%, according to Omedia.
NVIDIA GPUs are widely used in the AI and machine learning community due to their high performance and support for CUDA, a parallel computing platform and programming model for NVIDIA GPUs. As a result, OpenAI primarily uses NVIDIA GPUs (graphics processing units) for training and running its AI models, such as GPT-3 and ChatGPT.
CUDA is a parallel computing platform and programming model developed by NVIDIA for general-purpose computing on GPUs, making it useful for AI applications such as deep learning and machine learning. It allows for efficient and high-performance computations on large datasets. CUDA’s ecosystem of libraries, tools, and resources makes it easy for developers to take advantage of the GPU’s parallel computing capabilities, accelerate their applications and build more powerful and efficient AI models. It also includes libraries for deep learning, computer vision, and other AI workloads, as well as tools for profiling and debugging and resources for learning and training. Additionally, CUDA is widely supported by the industry, making it easy to integrate CUDA-enabled AI applications with other software and hardware systems.
While we feel confident that NVDA will be a big winner as a result of the AI arms race, nothing in investing is a sure thing.
We are worried that competition may arise as AI becomes increasingly significant and NVIDIA’s profits grow. Specifically, we are concerned about competition from China, a country with vast technical resources, a significant market for semiconductors, and a political desire to establish a domestic supply in light of the US government’s technology restrictions.
For example, Chinese GPU developer Moore Threads has raised $215.4 million in Series B funding to support its ongoing research and development of multi-functional GPUs. This funding indicates investors’ belief in the potential of Chinese GPU development, driven by the country’s need for AI/ML accelerators and graphics processors for gaming. Since its establishment in late 2020, Moore Threads has launched two graphics processors that are compatible with its MUSA computing platform: Sudi and Chunxiao, along with multiple add-in-boards built around that base design. These multi-functional GPUs are intended for various industries, including gaming, cloud computing, AI/ML, FP32 high-performance computing, and virtual desktop infrastructure. This broad market approach has enabled Moore Threads to attract investors from various backgrounds.
Another source of competition is the development of AI chips by tech giants themselves. For instance, Google has announced a processor specifically designed for improved imaging and machine learning capabilities. They have emphasized that the chip can efficiently perform tasks such as real-time language translation for captions, offline text-to-speech, image processing, and other machine learning-based functions, such as live translation and captions.
Financials & Valuation
After several years of strong growth, NVIDIA experienced a slowdown in its fiscal year 23 (ending January) due to challenging comparisons in video games and data centers from previous years and the decline of the cryptocurrency market. As a result, sales for FY23 are projected to increase only 0.1%, in contrast to a 61.4% increase the previous year. This stagnation in sales is largely caused by a projected 30% year-over-year decline in the gaming segment, which is the most affected by cryptocurrency.
The sudden decrease in growth led to an increase in operating expenses that surpassed revenue growth, resulting in a 14% decrease in operating margins to 33.3% in FY23. This has in turn caused earnings per share to be projected to drop by 26% in FY23.
Amid a slowing economic climate and uncertainty surrounding end market demand, consensus projections remain cautious. For FY24, consensus estimates project sales growth of only 8.3% and earnings per share to recover to $4.39, which is still below its FY22 EPS of $4.44. However, with the escalating competition among tech giants in the field of AI, and the reliance on NVIDIA’s GPUs for processing AI workloads, there is potential for NVIDIA to surpass these conservative estimates.
In January 2023, NVIDIA experienced a significant rally of 43%. While we believe the market’s recognition of the opportunity presented by ChatGPT may have played a role, we believe the broader shift towards riskier assets had a larger impact. During the same period, Tesla (TSLA) experienced a rally of 64% and Bitcoin (BTC-USD) saw a rally of 40%.
The recent rally in NVIDIA’s stock has led to a forward 12-month EPS multiple of 48x. However, considering the potential impact of ChatGPT on the company’s earnings, it is possible that NVIDIA could exceed consensus estimates. If ChatGPT accelerates the company’s projected results by one year, NVIDIA could see earnings per share of $5.56 in FY24. At this level, the stock would be trading at a more attractive multiple of 37.5x.
The advent of ChatGPT has sparked increased competition among major technology firms to develop and improve their AI capabilities. As a result, these companies are expected to ramp up their investments in NVIDIA’s GPUs to support their AI operations. While current forecasts predict a moderate improvement in performance, there is potential for NVIDIA to significantly surpass these projections.