NVIDIA's Evolution: From Gaming Graphics to AI Powerhouse
NVIDIA: Why is everyone talking about it?
Within the last 6 months, if you were to have even glanced at a financial news source, it would be hard to miss the name NVIDIA.
With a meteoric accession to the top of global equity markets in 2023-24, NVIDIA now sits as the fourth largest company in the world based on market capitalisation. The company is described by Goldman Sachs as ‘the most important stock on planet earth’, and that is with good reason.
Regarded as the ‘worlds engine for artificial intelligence’, NVIDIA’s chips power the complex calculations required in AI software. The recent growth in popularity of such software, i.e. ChatGPT, have ballooned demand for NVIDIA’s products and facilitated their rise to power.
To grasp their simply unbelievable recent performance, here are some statistics.
Since the beginning of 2023, NVIDIA tripled their revenue, grew their net profit eightfold and increased their stock price by close to 250%. Further, their market cap grew 530%, adding a trillion dollars worth in 8 months. To put that in perspective, this short period of growth is worth more than the entire Turkish economy. And finally, almost unbelievably, they are estimated to have 80% of AI computation market share.
The rest of this article aims to explore how NVIDIA got here and, more importantly, where they are going.
How did NVIDIA get here?
From its inception in 1993, NVIDIA set out to be a gaming company, producing graphics processing units (GPUs) that enhanced the quality of video games. However, their path to global AI dominance started almost 2 decades ago, when CEO Jen Hsun Huang unleashed the potential of NVIDIA creativity. By allowing graphic chips to work on non graphical applications, developers quickly realised that NVIDIA GPUs could be utilised in non gaming sectors.
These ‘non-gaming’ verticals now span across robotics, automotive applications, healthcare, finance and AI.
But it is still hard to understand why NVIDIA has dominated competitors, i.e. Intel, over the last few years. The answer goes back to 2020, where the crypto revolution gripped the mainstream public. The prospect of the third digital revolution had every retail and institutional investor flocking to digital currencies. Consequently, NVIDIA’s GPUs, of which are pivotal to mine crypto, were met with insatiable demand, propelling them above Intel to the top of the chip market.
Such vigorous demand then continued to grow until it met the next technology wave, Artificial intelligence. Once this new frontier was breached, NVIDIA then began to experience an unprecedented level of demand that had exploded since early 2023.
Under an economic lens, a supply side outlook would be prudent, and in NVIDIA’s case is interestingly almost as important as their blistering demand.
Over time GPUs have become smaller and smaller, fitting into smaller and sleeker laptops than the cumbersome computers of the past three decades. However, the pace at which manufacturers have been able to shrink these components have plateaued. These small GPUs bring additional challenges, where finding ways to safely package these pieces become increasingly complex.
Overarchingly, NVIDIA’s supply bottleneck works against them. Being unable to satisfy high level demand leaves extra revenue on the table. However, these issues aren’t unique to NVIDIA. Across the GPU industry, production is hamstrung, with Elon Musk stating that ‘GPUs are considerably harder to get than drugs.’
Interestingly though, some elements of low supply work in NVIDIA’s favour. Under basic microeconomic principles, low supply ultimately provides firms pricing power. Having this level of pricing power in a market where 80% of demand is already held massively contributes to NVIDIA'S bottom line. These factors contributed to NVIDAs healthy gross profit (+70%) and net profit (~50%) margins, levels simply tantalising to investors.
Now, most importantly, we must now consider where NVIDIA is headed.
Where is NVIDIA going?
To truly understand NVIDIA as a company, we must consider its future roadblocks and growth prospects.
In the near-term it is clear that NVIDIA will continue to do well. As companies fight for a slice of the AI software pie, NVIDIA will continue to rake in revenue as the backbone of the industry. When your products are delivered to leading firms, i.e. Cisco Systems, in armoured cars, value is undeniable.
However, after combing through a variety of articles and NVIDIA’s 10-K, there are a few recurring issues that could pose risk to NVIDIA's future.
Firstly, the cyclical risks that come with lofty valuations. The speed at which NVIDIA has toppled valuation milestones in the last year has been nothing short of blistering, but also poses risk. Concerns that a ‘follow-the-flock’ mentality have inflated the true value of NVIDIA’s share price have been raised as of late. Ark Investments Cathie Wood has suggested that expectations of NVIDIA are too high, making constant Wall Street beats both unattainable and unsustainable.
Next, the structural risks that come with ‘in house’ production. While in the near term, the majority of AI development relies on NVIDIA, this is slated to change in the future. With a large portion of AI developers being huge, cash rich companies, the R&D capabilities of these firms could reduce demand for NVIDIA GPUs in the future. With increasing development from firms such as Google, Meta and OpenAI, the individualised needs of each firm can be met more effectively. This puts NVIDIAs more general chips at risk in the future, but the time horizon for this is still unknown. In recent weeks, a notable development in the ‘in house’ production avenue arose, where OpenAI’s Sam Altman put forward the intention of raising $7 trillion (!!!) towards AI infrastructure.
Finally, the political risks that come with geopolitical tensions. Primarily, the US’s weakened relationship with China comes at a detriment to NVIDIA growth. With fears that stronger AI capabilities will enhance chineses military and cyber espionage potential, the US government sanctioned NVIDIA chips to China. While NVIDIA still managed to pivot and sell lower level chips that met government regulations last december, changes to government policy can still impact growth. Another foreign relationship pivotal to NVIDIA is that between the US and Taiwan. Taiwan has historically been a hub for chip manufacturing, and such is more important now than ever. The majority of NVIDIA Chip manufacturing is undertaken by the Taiwan Semiconductor Manufacturing Company (TSMC). This allows NVIDIA to divert more attention to its core competencies while they allow TSMC to deal with the complexities of chip manufacturing. While there isn't any friction between the US and Taiwan at the moment, maintaining their ‘robust unofficial relationship’ is key to the future success of NVIDIA. With NVIDIA deriving 56% of FY2024 revenue from overseas, long term developments in US foreign policy could have broad impacts on NVIDIAs future.
Ultimately, NVIDIA is a company key to the future of AI, and its future growth is dependent on its ability to elevate supply bottlenecks, sustain innovation and maintain its moats.
NVIDIA: Why is everyone talking about it?
Within the last 6 months, if you were to have even glanced at a financial news source, it would be hard to miss the name NVIDIA.
With a meteoric accession to the top of global equity markets in 2023-24, NVIDIA now sits as the fourth largest company in the world based on market capitalisation. The company is described by Goldman Sachs as ‘the most important stock on planet earth’, and that is with good reason.
Regarded as the ‘worlds engine for artificial intelligence’, NVIDIA’s chips power the complex calculations required in AI software. The recent growth in popularity of such software, i.e. ChatGPT, have ballooned demand for NVIDIA’s products and facilitated their rise to power.
To grasp their simply unbelievable recent performance, here are some statistics.
Since the beginning of 2023, NVIDIA tripled their revenue, grew their net profit eightfold and increased their stock price by close to 250%. Further, their market cap grew 530%, adding a trillion dollars worth in 8 months. To put that in perspective, this short period of growth is worth more than the entire Turkish economy. And finally, almost unbelievably, they are estimated to have 80% of AI computation market share.
The rest of this article aims to explore how NVIDIA got here and, more importantly, where they are going.
How did NVIDIA get here?
From its inception in 1993, NVIDIA set out to be a gaming company, producing graphics processing units (GPUs) that enhanced the quality of video games. However, their path to global AI dominance started almost 2 decades ago, when CEO Jen Hsun Huang unleashed the potential of NVIDIA creativity. By allowing graphic chips to work on non graphical applications, developers quickly realised that NVIDIA GPUs could be utilised in non gaming sectors.
These ‘non-gaming’ verticals now span across robotics, automotive applications, healthcare, finance and AI.
But it is still hard to understand why NVIDIA has dominated competitors, i.e. Intel, over the last few years. The answer goes back to 2020, where the crypto revolution gripped the mainstream public. The prospect of the third digital revolution had every retail and institutional investor flocking to digital currencies. Consequently, NVIDIA’s GPUs, of which are pivotal to mine crypto, were met with insatiable demand, propelling them above Intel to the top of the chip market.
Such vigorous demand then continued to grow until it met the next technology wave, Artificial intelligence. Once this new frontier was breached, NVIDIA then began to experience an unprecedented level of demand that had exploded since early 2023.
Under an economic lens, a supply side outlook would be prudent, and in NVIDIA’s case is interestingly almost as important as their blistering demand.
Over time GPUs have become smaller and smaller, fitting into smaller and sleeker laptops than the cumbersome computers of the past three decades. However, the pace at which manufacturers have been able to shrink these components have plateaued. These small GPUs bring additional challenges, where finding ways to safely package these pieces become increasingly complex.
Overarchingly, NVIDIA’s supply bottleneck works against them. Being unable to satisfy high level demand leaves extra revenue on the table. However, these issues aren’t unique to NVIDIA. Across the GPU industry, production is hamstrung, with Elon Musk stating that ‘GPUs are considerably harder to get than drugs.’
Interestingly though, some elements of low supply work in NVIDIA’s favour. Under basic microeconomic principles, low supply ultimately provides firms pricing power. Having this level of pricing power in a market where 80% of demand is already held massively contributes to NVIDIA'S bottom line. These factors contributed to NVIDAs healthy gross profit (+70%) and net profit (~50%) margins, levels simply tantalising to investors.
Now, most importantly, we must now consider where NVIDIA is headed.
Where is NVIDIA going?
To truly understand NVIDIA as a company, we must consider its future roadblocks and growth prospects.
In the near-term it is clear that NVIDIA will continue to do well. As companies fight for a slice of the AI software pie, NVIDIA will continue to rake in revenue as the backbone of the industry. When your products are delivered to leading firms, i.e. Cisco Systems, in armoured cars, value is undeniable.
However, after combing through a variety of articles and NVIDIA’s 10-K, there are a few recurring issues that could pose risk to NVIDIA's future.
Firstly, the cyclical risks that come with lofty valuations. The speed at which NVIDIA has toppled valuation milestones in the last year has been nothing short of blistering, but also poses risk. Concerns that a ‘follow-the-flock’ mentality have inflated the true value of NVIDIA’s share price have been raised as of late. Ark Investments Cathie Wood has suggested that expectations of NVIDIA are too high, making constant Wall Street beats both unattainable and unsustainable.
Next, the structural risks that come with ‘in house’ production. While in the near term, the majority of AI development relies on NVIDIA, this is slated to change in the future. With a large portion of AI developers being huge, cash rich companies, the R&D capabilities of these firms could reduce demand for NVIDIA GPUs in the future. With increasing development from firms such as Google, Meta and OpenAI, the individualised needs of each firm can be met more effectively. This puts NVIDIAs more general chips at risk in the future, but the time horizon for this is still unknown. In recent weeks, a notable development in the ‘in house’ production avenue arose, where OpenAI’s Sam Altman put forward the intention of raising $7 trillion (!!!) towards AI infrastructure.
Finally, the political risks that come with geopolitical tensions. Primarily, the US’s weakened relationship with China comes at a detriment to NVIDIA growth. With fears that stronger AI capabilities will enhance chineses military and cyber espionage potential, the US government sanctioned NVIDIA chips to China. While NVIDIA still managed to pivot and sell lower level chips that met government regulations last december, changes to government policy can still impact growth. Another foreign relationship pivotal to NVIDIA is that between the US and Taiwan. Taiwan has historically been a hub for chip manufacturing, and such is more important now than ever. The majority of NVIDIA Chip manufacturing is undertaken by the Taiwan Semiconductor Manufacturing Company (TSMC). This allows NVIDIA to divert more attention to its core competencies while they allow TSMC to deal with the complexities of chip manufacturing. While there isn't any friction between the US and Taiwan at the moment, maintaining their ‘robust unofficial relationship’ is key to the future success of NVIDIA. With NVIDIA deriving 56% of FY2024 revenue from overseas, long term developments in US foreign policy could have broad impacts on NVIDIAs future.
Ultimately, NVIDIA is a company key to the future of AI, and its future growth is dependent on its ability to elevate supply bottlenecks, sustain innovation and maintain its moats.
