1.

Indentify whether your company has a competitive advantage or disadvantage in its

primary industry. Its primary industry is the one in which it has most sales.
Founded in 1993, Nvidia Corporation has grown from a company specializing in graphics
processing units (GPUs) for gaming to a giant in the semiconductor industry. Currently, with the
fastest and most miraculous growth segment thanks to the explosion of AI and the need for big
data, accounting for 70% of the company's revenue, Data Center & AI has become Nvidia's core
area. Especially in the field of AI and Deep Learning, Nvidia has held over 95% of the GPU
market share serving this field, when large technology corporations with a commitment to invest
325 billion USD in AI infrastructure in 2025 such as OpenAI, Microsoft, Google, Meta or
Amazon are all strategic partners of Nvidia when using GPUs for AI produced by this company.
Meanwhile, AMD only holds 10% of the GPU market share and focuses mainly on gaming, Intel
has made efforts to invest in AI GPUs with Ponte, Vecchio, Gaudi but still cannot compete
completely with Nvidia in terms of performance and software. Besides, in the gaming market,
Nvidia holds about 80% of the global gaming GPU market share. From there, we can see
Nvidia's market dominance in providing GPUs for gaming and AI.

Nvidia currently holds many competitive advantages in major markets. First, Nvidia has always
invested heavily in R&D, with the cost of this segment being $8.1 billion, accounting for 22% of
the company's revenue, far exceeding other competitors with 17% of AMD, 14% of Intel. This
move reflects the company's focus on quality and innovation in its products, ensuring that the
company is always leading in its main products in the market. Along with that is the competitive
advantage in products, with parallel processing capabilities and advanced chip architecture.
Specifically, Nvidia GPUs can process thousands of calculations at once, while CPUs only
process sequentially, RTX 4090 has up to 16,384 CUDA cores, helping to process graphics, AI
and physics simulation faster, and H100 AI GPU has up to 80 billion transistors, providing
performance up to 30 times faster than CPUs in AI training. In addition, using the 3nm process
from TSMC has helped Nvida's products increase performance but save energy, and also support
PCle 5.0, NVLink allows data transmission 4 times faster than the old generation GPU. Those
are huge competitive advantages in terms of products that Nvidia creates compared to its
competitors. Not only that, Nvidia also monopolizes AI software and its CUDA development
ecosystem. CUDA is not only a software platform, but also an ecosystem that locks in customers,
creating a huge barrier for competitors. With support for TensorFlow, PyTorch, JAX, more than
4 million developers are using CUDA for AI, scientific simulation and deep learning. That is
why Nvdia's GPUs are irreplaceable in AI training, making the cost of switching customers to
AMD or Intel very high because they have to rewrite the entire code, not to mention, AMD and
Intel do not have a strong ecosystem like CUDA. Next is the ability to expand into cloud
computing and AI services when Nvidia is currently providing AI services on the cloud through
DGX Cloud, helping their customers run AI models without having to buy GPUs. Nvidia also
cooperates with Microsoft, Google Cloud, Amazon AWS to provide GPU-as-a-service. From
there, we can see that Nvidia is transforming from a hardware company to a comprehensive AI
service provider, creating a huge competitive advantage in the global AI market. Along with the
advantage in service products, supply chain and manufacturing strategy is also an advantage
when cooperating with TSMC, a manufacturing company with the most advanced technology
processes, ensuring high performance and energy efficiency for advanced chip products. Thanks
to this cooperation, Nvidia does not need to directly spend a large amount of fixed costs on
building factories and manufacturing infrastructure but still has high-tech products to meet
customers. In addition to TSMC, Samsung is also a trusted manufacturing partner of Nvidia, and
Nvidia is also actively cooperating extensively with Southeast Asian countries to have more
high-quality but low-cost production sources. In terms of valuation and profit, Nvidia is currently
overwhelming with terrifying numbers with a forecast of AI revenue of 100 billion USD in 2026,
and Nvidia's company valuation is currently 4 times that of rival semiconductor companies such
as AMD and Intel combined.

However, Nvidia faces many challenges in the AI-related product segment. In a lucrative pie,
large and small companies around the world cannot sit still. Specifically, Google has launched
Google TPU, which is said to be more effective than Nvidia GPUs in AI inference. AI startups
such as Groq, Cerebras, Tenstorrent are developing dedicated AI chips to reduce dependence on
Nvidia GPUs. Not only that, in the international market, the rise of DeepSeek in China, with an
AI model developed with high performance but 40% lower cost than Nvidia. And the Chinese
government and large corporations such as Alibaba, Baidu, Huawei are all investing heavily to
produce domestic AI chips. From there, we see fierce competition coming from everywhere. Not
only that, if domestic Chinese companies can produce AI chips themselves, Nvidia will lose a
huge market share here, which has contributed a significant part to Nvidia's revenue. Not to
mention a major disadvantage for Nvidia is political instability, as tensions escalate between the
US and Chinese governments. It is the US's restrictions on AI chip exports to China that have
caused Nvidia's revenue to drop by up to 30% in 2024 in this market. Not to mention that TSMC
is a factory in Taiwan, which is also heavily dependent on this conflict. Therefore, in the future,
unforeseen risks may occur and affect Nvidia's supply chain, if the company does not have
timely solutions.

2. Evaluate your company against the four generic building blocks of competitive
advantage: efficiency, quality, innovation, and responsiveness to customers. How does this
exercise help you understand the performance of your company relative to its competitors?
We will analyze each core element of competitive advantage. First is performance. Nvidia
always optimizes its production and supply chain, typically applying the Fabless production
model. The company does not produce chips itself, but instead reduces production effort and
resources by cooperating with manufacturing companies around the world, specifically TSMC
and Samsung to launch new generation AI GPUs using 3nm and 5nm processes. In addition,
Nvidia is taking specific actions to reduce geopolitical risks such as cooperating with more
factories around the world, or TSMC itself investing 40 billion USD in expanding production in
Arizona. Next is a very effective inventory management policy. Nvidia has applied the Just-in-
Time (JIT) inventory strategy, which helps to significantly reduce inventory costs and increase
the ability to respond to unexpected market fluctuations. The above factors have helped Nvida
achieve a profit margin of 65%, higher than its two competitors AMD and Intel with 50% and
40% respectively.

Next is product quality. Nvidia's GPUs are very durable and efficient. The H100 Tensor
Core GPU is currently the world's most powerful AI GPU, providing 30x performance
compared to CPUs in AI training. Meanwhile, NVLink data transmission speed is 4x faster
than PCle 5.0, helping to optimize AI data processing speed. The RTX 5000 Series
improves Ray Tracing and DLSS 3.5, helping to enhance the gaming and AI experience for
users. Along with that, Nvidia is ISO 9001 certified and has received a "letter of
confirmation" for the IATF 16949 automotive quality standard, as a milestone affirming
product quality. In terms of product performance when training AI, Nvidia's H100 is rated
at 100% with 700W, AMD's MI300X is 80% with 600W, Google TPU v5 is 75% with
550W, and Intel Gaudi 3 is 65% with 450W. This shows how superior Nvidia's products
are in terms of AI GPU and gaming.

Next is innovation, Nvidia is famous for its strong investment in R&D with 8.1 billion USD
equivalent to 22% of revenue in 2025, higher than AMD and Intel with 17% and 14%
respectively. Important projects that Nvidia is aiming for include Quantum AI Computing,
Omniverse Digital Twins, AI Cloud GPU. These are all projects with technology that has
the potential to lead in the future, when Quantum AI Computing is truly the future door of
AI, and AI Cloud is the expansion of necessary AI services, diversifying its ecosystem.
Nvidia has now partnered with AWS, Microsoft Azure, Google Cloud to provide its GPUs
as a subscription service, aiming to expand the company's AI Cloud project. From there,
we can see that Nvidia is leading in innovation in the market, with a huge amount of
investment and effort.

Finally, when it comes to software support and customer service, Nvidia is rated by
Comparably with a customer satisfaction score (CSAT) of 83/100, with 51% of customers
being “Very Satisfied” and 32% “Satisfied”. Looking deeper, Nvidia has provided
customer support through a variety of channels, including online support, public forums,
and a customer service center, with the aim of meeting the needs of both individual and
enterprise customers. In addition, Nvidia has deployed AI solutions to improve customer
service, including using virtual assistants and intelligent chatbots to resolve customer
requests efficiently and quickly. On a macro level, Nvidia has made a big step forward with
AI Cloud, allowing customers to access GPUs remotely, instead of having to buy hardware.
Along with that, Nvidia also showed its ability to meet customer needs and adapt quickly
when it launched GPU versions with reduced performance (H800, A800) to help its
customers in China solve the problem of the US export ban.

In summary, Nvidia has strong advantages in all four competitive factors, but there are still
challenges from strong competition from large and small companies around the world, and
geopolitical risks affecting the company's business and supply chain.

3. What are the distinctive competencies of your company?

For Nvidia, the following unique capabilities have helped the company maintain its
leadership in the semiconductor and AI industries.

One of Nvidia's most important capabilities is the CUDA platform. CUDA supports all
major AI frameworks, including TensorFlow, PyTorch, JAX, which is now the standard in
AI training, making Nvidia GPUs almost irreplaceable in AI training when more than 90%
of the world's AI data centers and 4 million developers are using CUDA software to
optimize GPUs. Not only that, Nvidia is currently integrating CUDA with new platforms
such as Quantum Computing and AI Cloud to expand the scope of applications.
Competitors in the market such as AMD with ROCm or Intel with oneAPI have not
achieved the same level of popularity as CUDA, and if a business wants to switch from
Nvidia GPUs to AMD or Intel, they have to rewrite all the AI code, which is very resource-
intensive and labor-intensive. From there we can see that CUDA is not only a software but
also a strategic advantage that helps Nvidia control the AI market.

Next is the huge R&D investment budget. As mentioned in the previous section, with a huge
R&D budget, Nvidia always launches new, advanced products leading the market in AI GPUs,
specifically the H100 and B100. Currently, Nvidia is developing a new generation of AI GPUs
such as the H200 to optimize AI training and inference, along with cooperating with IBM to
develop Quantum AI Computing to create more market-leading products in the future. The AI
super chip "Grace Hopper" - the first CPU-GPU combination chip to optimize AI training and
HPC, along with billions of dollars invested in the DGX Cloud platform, helping Nvidia develop
AI Cloud Computing, are also 2 important R&D projects that Nvidia is implementing. From
serious and huge investments, Nvidia always aspires to be a market leader, a pioneer and explore
the limits of technology and engineering, making Nvidia always more special than its
competitors.

And thanks to the large investment, Nvidia's AI ecosystem is increasingly diverse. In addition to
physical GPUs and AI Cloud, Nvidia also invested in developing Nvidia Earth-2 Supercomputer
to simulate climate change and complex AI research, or Nvidia DRIVE Orin&Atlan, helping
self-driving cars process AI faster, Isaac Robotics Platform helps apply AI in manufacturing and
logistics. From there, we can see that the AI ecosystem that Nvidia creates is very diverse,
creating a difference and uniqueness for Nvidia.

Nvidia not only relies on its own resources to develop, but also cooperates with many large
technology corporations in the world such as Google, Microsoft, Amazon or Meta to ensure a
solid, stable and increasingly expanding position in the market. These major partners have
invested more than 325 billion USD in AI infrastructure, helping Nvidia benefit greatly.

And finally, we cannot fail to mention Nvidia's flexible manufacturing model - Fabless strategy.
Although there are many risks when affected by political fluctuations, Nvidia has a solution to
diversify its supply chain. And thanks to this model, Nvidia has reduced costs and is more
flexible in production than its competitors, helping the company focus resources on design and
R&D, helping to bring the best products to market.

And with these unique capabilities, it is no surprise that Nvidia remains the market leader. The
above factors have helped Nvidia create great value for the company, rarity in the market, not
easy to copy, helping to maintain a long-term competitive advantage, and are well-organized to
exploit for sustainable development. With these advantages, Nvidia is not only the leader but
also sets the standard in the AI and GPU industry, and is expected to continue to lead the market
until at least 2030.
5. Do the strategies your company is pursuing now build on its distinctive competencies?
Are they an attempt to build new competencies?

Nvidia's current strategy is closely linked to its unique capabilities. First, they always pay
attention to expanding the CUDA software ecosystem, which is the key for Nvidia to build an
ecosystem and retain customers. Because they know that the switching cost of customers leaving
CUDA is currently high, Nvidia has a strategy to continuously improve and upgrade CUDA
services to retain customers and expand the market, creating barriers to entry for competitors.
Next, which is both a core capability and a unique strategy, is R&D investment projects. That is
one of the core factors that helps Nvidia have market-leading products. Nvidia has a strategy to
launch many new research projects such as Blackwell GPUs, Quantum AI Computing, super
chips and AI Cloud. Not only that, with the strategy of expanding the business model and
market, that has been demonstrated when Nvidia has taken advantage of the company's
cooperative relationships with large technology corporations, which are inherently customers in
the AI field, to jointly invest in developing new product projects, thereby controlling the
company's position in the future AI market. In terms of production capacity, Nvidia currently has
a strategy to expand factories in new markets in Southeast Asia, to minimize political risks and
increase production capacity. Through that, we can see that Nvidia's development strategy is
closely linked to the company's core competencies, increasingly developing its core strengths,
along with expanding its capacity, market and products.

6. What are the barriers to imitating the distinctive competencies of your company?

Nvidia's leading position is a position that many companies want to have, but there are
many barriers for companies to imitate and achieve the capabilities that have helped
Nvidia achieve its current position. First is the technological barrier. Nvidia's current
success is greatly contributed by core technology, superior design and being ahead of
competitors, making them always the market leader for a long time. Specifically, CUDA, a
proprietary software ecosystem, the world's most powerful parallel programming
platform, trusted by more than 90% of the world's AI data centers, more than 4 million
developers. Not only that, Nvidia is increasingly developing CUDA. In fact, this is a huge
barrier for other companies to copy, not to mention the very high cost for Nvidia's
customers to switch away from using CUDA. Another example of the technology barrier is
the high-performance AI GPUs with H100, B100, which have 20-30% higher AI
performance than competitors. Currently, there are some competitors on the market that
have advantages in AI inference tasks, but they cannot compete and completely copy
Nvidia's GPUs in AI training. And to keep up with the technology that Nvidia currently
has is also costly in the billions of dollars and takes a long time.

Next is the financial barrier. We can see that thanks to the early technological
breakthrough, Nvidia has brought a large number of customers, orders, and of course huge
revenue and profits. From there, Nvidia has been even more aggressive in researching and
developing new technologies. With spending of up to $8.1 billion on R&D, it is a huge
number that very few corporations dare to invest in, especially new startups. Nvidia is also
aggressively expanding projects related to AI Cloud and Quantum AI Computing to
continue leading the AI market in the future. Because of this loop, Nvidia is creating a large
barrier and gap with rival companies in terms of technology and revenue from the AI
market.

Another barrier is manufacturing and supply chain. Currently, TSMC's 3nm manufacturing
technology has no competitor to catch up with. Competitors like Intel are still having difficulty
producing chips under 5nm. Not only that, after having had success in cooperating with TSMC
and Samsung for production, Nvidia has been proactively cooperating widely with factories in
Southeast Asia to expand its supply chain, making them always proactive in the market, making
it difficult for competitors to keep up.
Not only that, Nvidia's strategic partnerships with bigtech are very broad. They cooperate with
Amazon, Google, IBM and a number of other companies. Nvidia also invests in AI startups, to
ensure that they always receive the first breakthrough ideas. This makes Nvidia very diverse and
is a big barrier for competitors who want to penetrate and compete.
Finally, there are barriers in terms of law, export control and intellectual property rights. Nvidia
has more than 15,000 patents in the field of AI & GPU, making it impossible for competitors to
copy its technology. Along with that, bans on exporting Nvidia's most advanced super chips to
China also make it difficult for Chinese companies to research Nvidia's most advanced
technologies.

7. Is there any evidence that your company finds it difficult to adapt to changing industry
conditions? If so, why do you think this is the case?
The semiconductor industry is a competitive and rapidly changing industry, and is affected by
many factors. Let's analyze the challenges from changes in the industry and how Nvidia is
dealing with them. First, there are changes related to the US's export control policy towards
China. This has caused Nvidia to lose a large amount of revenue from customers in China, which
contributes a large amount of revenue to Nvidia. However, Nvidia has adapted by producing
lower-performance versions of its chips for export to China, but it is still significantly limited.
Although the company has a solution to solve the revenue problem, the heavy dependence on the
Chinese market is still a problem that has no effective alternative solution.
Another aspect is the difficulty in the supply chain and manufacturing. Nvidia has been very
successful with TSMC and Samsung to produce AI chips and achieve high profit margins.
However, due to the sudden increase in demand for AI chips, along with the scarcity of raw
materials due to political tensions in China and Russia, the shortage of high-quality human
resources has created challenges for the semiconductor industry in general and Nvidia in
particular. Not only that, TSMC is a Taiwanese company, very vulnerable to US-China political
tensions. Therefore, Nvidia's current production capacity is very vulnerable. To deal with that,
Nvidia has actively cooperated with places around the world, especially Southeast Asia,
developing countries to build factories, to minimize production risks, along with training high-
quality human resources but at low prices. Not only that, TSMC also plans to build a factory in
Arizona to help Nvidia reduce concerns about political tensions. However, in terms of raw
material sourcing, it is still a challenge for Nvidia and needs solutions to be able to solve it in the
future.
Next is the fierce competition of rivals every day and every hour. The rise of Intel, AMD,
Google TPU or startups like Groq, Tenstorrent, DeepSeek makes Nvidia's position replaceable at
any time. Especially in the AI inference segment, competitors are showing signs of products with
the same or better performance as Nvidia, but at a cheaper price. In response to this extremely
competitive market, Nvidia has invested billions of dollars in R&D to constantly develop its
products, and invest in many leading projects in the future, however, it still needs to change and
develop faster to always ensure its unique position.
Therefore, we see that the semiconductor market changes very quickly and Nvidia has also made
moves to adapt to the changes, but everything is still in the initial stage and the process will
encounter certain difficulties, especially in dealing with political conflicts and developing
products to always maintain the number one position in the market comprehensively.
References:

1. Semiconductor Industry Association. (2024). State of the U.S. Semiconductor Industry

Report 2024.
2. Sourceability. (2025). The Semiconductor Market Going into 2025: A Landscape of
Opportunities and Challenges.
3. Statista. (2025). Artificial Intelligence - Global Market Forecast.
4. Deloitte. (2025). 2025 Semiconductor Industry Outlook.
5. J.P. Morgan. (2025). AI Investment Trends 2025.
6. Harding Loevner. (2024, April 23). NVIDIA's Competitive Structure May Be More
Fragile Than Its Valuation Implies. Harding Loevner.
7. NVIDIA Developer. (2023, July 25). Advancing Humanoid Robot Sight and Skill
Development with NVIDIA Project GR00T. NVIDIA Developer Blog.
8. Cascade. (2022, November 25). How NVIDIA dominated the graphics processing space
with its perfect strategy. Cascade Strategy.
9. DCF Modeling. (2025, January). NVIDIA Corporation (NVDA): SWOT Analysis. DCF
Modeling.
10. The Strategy Story. (2023, March 15). NVIDIA SWOT Analysis. The Strategy Story.
11. Investor's Business Daily. (2024, August 30). How Nvidia Is Building A Competitive
Moat To Fend Off AI Challengers. Investor's Business Daily.
12. The Wall Street Journal. (2024, August 15). Why Every Big Tech Company Has Failed to
Dethrone Nvidia as King of AI. The Wall Street Journal.
13. VnEconomy. (2025, January 15). Nvidia đối mặt nguy cơ doanh thu giảm mạnh do lệnh
hạn chế xuất khẩu chip AI của Mỹ. VnEconomy.
14. Stockbiz. (2024, December). Nvidia thống trị năm 2024, nhưng năm tới sẽ gặp nhiều
thách thức?. Stockbiz.
15. VnExpress. (2022, February 8). Lý do vụ sáp nhập Nvidia - ARM sụp đổ. VnExpress.
16. Investors. (2025). Nvidia’s Cloud AI Strategy: How It Competes with AWS, Microsoft,
and Google Cloud. Investors.com.
17. FT. (2025). Nvidia Expands Into Robotics with Jetson Thor AI Chip. Financial Times.
18. Diễn Đàn Doanh Nghiệp. (2025). Toyota Hợp Tác Với Nvidia Trong AI Xe Tự Lái.
Diendandoanhnghiep.vn.
19. Nghiên Cứu Chiến Lược. (2025). Nvidia Đầu Tư R&D Tại Việt Nam - Hướng Đi Chiến
Lược Mới. Nghiencuuchienluoc.org.