I am a firm believer in tangible infrastructures.
If a company is in a business model of creating a physical product, that can deliver a certain function that is hard for its competitors to replicate, then this company is worth spending time to study it deeper.
On the flip side, companies that create physical products or infrastructures are usually asset intensive to a certain extent. They usually have a lower profit margin as well.
What’s even worse, is an asset-intensive company delivers a product or service with little no to competitive advantage over its peers.
Telcos and property development companies come to mind.
However, on the other end of the spectrum, is where we see examples of companies selling tangible products and are yet still very profitable. Apple is an example since it’s still a hardware-dominant company. TSMC, a company that spent billions in capital expenditures building new foundries, is a company that commands an above-average profit margin.
But none stole more thunder than NVIDIA Corporation (NASDAQ: NVDA). Why did this GPU maker rally so much?
Should you or should you not buy NVIDIA right now?
GPU is the infrastructure for AI development
For those who are not familiar with or clueless about chips, here is a simple-to-understand course.
There are mainly 3 types of chips, and a computer usually cannot work without either one missing. They are:
- Central Processing Unit (CPU): The mainframe and brain of a computer. CPUs are responsible for carrying out instructions and data processing.
- Graphics Processing Unit (GPU): GPUs are designed to accelerate graphics processing. Anything that has to do with visuals, color, and gaming, would require a GPU. And GPUs have also been the basic pillar of artificial intelligence.
- Memory chips: There are two main types of memory chips: random access memory (RAM) and read-only memory (ROM). RAM is used to store data that is being actively used by the computer. It is much faster than other types of memory, such as hard drives and solid-state drives. ROM is used to store data that is not changed very often, such as the computer’s operating system. It is much slower than RAM, but it is also much more durable.
Although most of the AI experiences that we are experiencing right now are mostly all software and intangible, GPU is a key infrastructure that AI cannot do without.
Which company will emerge triumphant in the battle for supremacy in AI is still too early to tell.
But in terms of a dominant GPU company, NVIDIA is one of the largest names in the segment.
Nvidia GPUs – Not the only ones used for AI and LLM. But currently the best
Nvidia is not the only GPU company. So it is not the sole beneficiary when it comes to providing key infrastructures for AI advancement.
But as of now, it is the best in class.
Nvidia’s A100 chips are currently the backbone of most, if not all of deep-learning AI.
AI hardware remains strongly consolidated to NVIDIA as of now, as reported by State of AI Compute report.
NVIDIA is no longer a pure gaming company
One major difference I noticed when comparing NVIDIA against Intel Corporation (NASDAQ: INTC) and Advanced Micro Devices, Inc. (NASDAQ: AMD) is that NVIDIA has pivoted over the years:
Data center became the core component within the span of 4 years.
AMD has its strength, which is in PC market processor shipments. But this strength turned into a major drawdown as AMD suffered from a weak PC market and inventory correction in the PC supply chain.
Even though NVIDIA’s gaming segment also got affected, its data center business remained strong.
CUDA – NVIDIA’s answer to rule them all
CUDA, or Compute Unified Device Architecture, is a parallel computing platform and application programming interface (API) that allows software to use certain types of graphics processing units (GPUs) for general-purpose processing.
NVIDIA developed CUDA way back in 2007. Today, it is used by millions of developers around the world. Although CUDA is not the only platform for GPU computing, it is the most mature in the space.
CUDA software development is available for free, but it still does require a compatible GPU. And that is where the business case of NVIDIA gets justified.
One thing though – NVIDIA is trading at a high premium now
Price is what you pay, value is what you get.
The hypergrowth mentality of buying companies at ridiculous valuations seems to creep in if you think of how the company might benefit or even thrive. But when the valuation comes to play, some people would want to think twice.
NVIDIA is close to its historical high. It is trading at a Price to Sales ratio of 25.15x and a Price to Earnings ratio of 67x.
In such an uncertain climate, even if the thesis and analysis play out well, it’s more likely to be a game plan for the next few years.
The uncertainty of the economic climate, should things turn for the worst, could see NVIDIA prices tumbling down.
NVIDIA: To buy or not to buy?
I’d be honest, I wouldn’t buy NVIDIA at the current price in the current situation.
Yes, I still pride myself to be a long-term growth investor. I like how much advantage it holds over its competitors.
But that does not mean we should buy irresponsibly without taking into consideration of the potential drawdowns.
NVIDIA might be a great business. But great businesses are not spared when a recession happens.
It is a waiting game now for valuation to normalize, or a crash to come.
Their latest Q1’23 was better than analysts’ expectations. But what really blew the price to dizzying highs is the guidance for Q2’23. The company forecasts revenues of USD 11 billion, which is a mind-boggling 52% ahead of the USD 7.1 billion consensus revenue estimate for the quarter.
I know. It is not every day that companies can report guidances that surpasses estimates. But as the superior guidance is now out of the box, the share prices of Nvidia are now in frenzy mode.
What do you think? Is your FOMO taking over you, or are the facts and figures laid out in this article much more convincing?