How AMD Could Benefit From "Hyperscalers Want Choice"

There has been a lot of news and article on Nvidia already, but one thing we need to take note of AMD is that the stock should get a boost from its AI push in 2024. The MI300 could reach $2 billion in annual run rate revenue by the end of the year.

And as an investors I would consider AMD as its stock is cheap compared to Nvidia. In this article, I would like to share what I have found out on what advantage AMD has as number 2 as AI chip customers are looking to keep their prices down while some might have embark on their own chips, but demand for cloud service providers and some enterprises would still be there and still growing.

AI chip customers, such as cloud service providers and enterprises, appreciate having a strong No. 2 in the market to keep prices down and spur innovation.

These hyperscalers want choice, they do not want to be tied down by a single supplier, because from business perspective, the company will be held ransom on the pricing if that happen.

So with AMD increasing their presence in the AI accelerator market late last year. This has prompted Nvidia to change their cycle for new product launches from two-year to each year. We have seen Nvidia new flagship product each year to match AMD's annual cadence.

One more factor that worth observing is the backlog created because chip makers want to pump up their sales, Nvidia have been selling every H100 GPU it can make and has a sizable backlog.

Nvidia has shown its inability to meet demand for AI chips, that has created opportunities for other chip makers to gain market share. If we look at the over market, by all accounts, the market for AI chips in data centers is massive and can support numerous vendors.

While Nvidia has some improved product, like H200 lined up to meet demand, AMD has a new data center GPU, the MI400, in its near-term product pipeline.

The most recent news on last Friday (17 May 2024), we saw Microsoft announcing that they will offer their cloud customers AMD alternative to Nvidia AI processors, we will need to see more details from Microsoft.

As we move into a different phase of AI, and we could say it is maturing stage for generative AI applications. I foresee there will be no slowdown in spending for data center infrastructure, these are essentials to support these applications.

We have already seen in late 2022, Startup OpenAI started the generative AI trend with its buzzy chatbot, ChatGPT. It has since evolved and there are many other large language models which is the primary important piece of technology.

While Nvidia could provide a full-stack solution for AI computing, including its CUDA software and foundational models and libraries to jump-start development of AI applications, but the way things are moving, I would think companies are looking for value and performance.

AMD is trying to compete with different offering of value and performance, though no full stack of software and predefined libraries, these are not of any concerns as of now. In the initial stage of generative AI, that might be relevant.

These platforms will offer its customers an alternative to Nvidia’s H100 family of powerful graphics processing units (GPUs) that currently dominate the AI chip market for data centres. However, customers find it hard to obtain Nvidia’s chips due to their high demand.

While AMD is upbeat about where the market is headed. Their CEO has said on 30 Jan in a conference call with analysts that the accelerator market could grow to about $400 billion by 2027. That figure took Wall Street by surprise because it was well above their forecasts.

Despite this, I feel that we need to look at how AMD has performed as compared to Nvidia and whether AMD does show potential of long term upside.

If we looked at how AMD stock price have performed compared to NVDA, there is some relationship, while any large movement in NVDA, could affect AMD stock price.

This might change after NVDA earnings on 22 May 2024 after market, because if there is any sign that Nvidia is losing market share and their challenge on backlog is still pending, we might see AMD present an opportunity.

If we looked at the current AMD traded price, it is above the 50-day period, but KDJ J value seem to suggest that it is forming a SELL signal.

This might be due to investors who has bought on either 22 April or 13 May, we need to watch the Nvidia earnings on 22 May as well, this could show us some movement to AMD.

I would hold on to load up AMD as there might be some price adjustment, maybe after 22 May would be a better timing.

While the spending on AI is increasing, but companies are looking for alternatives so that they can better manage their cost expenses, as we have seen some of the earnings report that cost expenditure is something that investors look out for.

AMD provided a choice and alternatives to organizations to spend on AI, and this could help to make their strategy more cost efficient. So I personally would think that it is good to monitor AMD and look out for a good time to load more shares.

Appreciate if you could share your thoughts in the comment section whether you think AMD stock price would benefit from its alternative role for AI chips choice?

Disclaimer: The analysis and result presented does not recommend or suggest any investing in the said stock. This is purely for Analysis.