By Hashmeel Suka
Nvidia designs and develops hardware processing units, supplying a large and diverse customer base spanning both the computing and the automotive industry. The company was founded in 1993 and has continued to operate as a fabless chipmaker, meaning it specialises only in the design and selling of semiconductors (silicon-chips), whereas actual manufacturing is outsourced to thirdparty foundries such as the Taiwan Semiconductor Manufacturing Company (TSMC) and Samsung.
The company's initial focus was central processing units (CPUs), which could perform basic computing tasks. Rapid advancements in technology, however, meant that graphical elements would become too demanding for the hardware available at the time. In response, Nvidia pioneered the graphics processing unit (GPU), which significantly improved the rendering of multidimensional images and video, used in media and gaming. The utility and importance of GPUs (due to being adept at processing large computational tasks at high speeds) has itself grown rapidly, with the technology now being applied to other disciplines such as machine learning (ML), artificial intelligence (AI) and cryptocurrency mining.
In addition to CPUs and GPUs, Nvidia is involved in the development of data processing units (DPUs), which use network hardware to efficiently share and transfer large amounts of data across data centres. DPUs are particularly important for cloud computing providers such as Google, Microsoft Azure, and Oracle.
Operating structure
Nvidia operates across two main segments namely Compute & Networking (~55% of total revenue) and Graphics (~45% of revenue). The Compute & Networking segment has quickly become the main area of growth for the company and includes the Data Center, Automotive and AI-related businesses. The Graphics segment is focused primarily on the gaming offering, which is famous for the GeForce GPU product range, as well as the GeForce NOW game streaming service.
Financials
Growth over the past five years has been quite strong. Since FY19, Nvidia has achieved revenue and earnings growth (on a compounded annual basis) of ~39% and ~51%, respectively. This is particularly impressive considering the impact of the Covid-19 pandemic, geopolitical tensions, and macroeconomic uncertainty throughout this period.
FY23 marked a slowdown in growth for the chip developer:
In FY24, the group saw an exceptionally strong rebound, with demand underpinned by the advent of ChatGTP and other AI applications:
November 2022 saw the launch of ChatGPT (GPT-3.5) by US-based research organisation, OpenAI. Along with it came an extensive push toward AI, with both customers and competitors racing to get a grip on the "revolutionary" technology.
This was particularly encouraging for Nvidia, being one of the world's largest semiconductor companies. At the time, revenue from the Data Centre business (which reports sales of chip-hardware used in AI functionality) was estimated at ~$10.5 billion, while the total addressable market for enterprise AI was forecasted at ~$100 billion. As of early 2024, Data Centre revenue was ~$47.5 billion, with the total addressable market now believed to be worth a whopping $310 billion, as prospects within AI have improved and remain immense.
It has, however, not been all smooth sailing for the company. From mid-2022 through late-2023, the US government placed various export restrictions on the sale of high-end chips and supercomputers, particularly to China. The Chinese market (including Hong Kong) accounts for between 20% and 25% of Nvidia's total revenue and this is spread across all key revenue segments of the company. The impact of these restrictions has been somewhat felt by the company- reports show that the top line took a knock of more than $400 million in FY23 as a result.
Lost revenue though is not such a major concern. In fact, many Chinese companies reportedly increased their order quantities in anticipation of the ban, thus ultimately softening the impact to Nvidia. Other workarounds were also employed such as the rechannelling of products though other regions; though these measures have since been impeded upon. Ultimately, these export restrictions may limit growth. Although demand from the US and rest of the world is set to remain strong, Nvidia will forego any surplus growth from China, where the AI market is expected to develop rapidly.
Investment case
Risks
Consensus outlook and valuation
Bloomberg consensus is extremely positive on Nvidia, with around 90% of sell-side analysts maintaining a "buy" recommendation on the stock. The 12-month aggregate target price is currently ~$995, which is ~12% above current levels (~$888).
Nvidia is trading on a 12-month blended forward PE ratio of 34 times (which is quite attractive compared to its long-term history) with the market still pricing-in significant earnings growth to come. Compared to other chip developers, the company trades at a premium of ~9%, reflecting a strong contraction over the past six months (five-year average premium: 62%).