Is AMD being overlooked?

I haven't been paying any attention to AMD and I just assumed that it would be a perpetual second place to NVIDIA. However, I recently came across Antonio Linares' content on Twitter and Youtube, and apparently he thinks I don't know my stuff.

Here is a recent video from him speaking about AMD.

Linares believes that trust and customer relationships are as crucial as technology in gaining market share in AI and AMD's strategy of patient market penetration has proven successful in CPUs and may work for AI GPUs and that AMD's long-term success may depend on its ability to expand beyond data centers into edge AI (inference rather than training).

Currently, the market is skeptical about AMD's prospects in the long term, and its stock is at about 25% down from its previous high. However, that presents an opportunity for long-term investors in the industry.

Summarizing with Fabric

I used fabric to summarize the video. I have written about Fabric before and highly recommend installing and using it in your knowledge workflow

yt https://www.youtube.com/watch\?v\=DZbO0QAgJJo | summarize  

One Sentence Summary:
AMD is poised to become an end-to-end AI giant, with significant progress in data center, client, and embedded segments despite market skepticism.

Main Points:

1. AMD revised AI GPU sales guidance for 2024 to $5 billion, up from $2 billion initially.
2. Data center revenue now accounts for 52% of total revenue, showing exponential growth.
3. Meta deployed 1.5 million EPYC CPUs and broadly adopted MI300X for inferencing infrastructure.
4. Microsoft is using MI300 for multiple co-pilot services powered by GPT-4 family models.
5. AMD's operating margin in the data center segment increased from 19% to 29% year-over-year.
6. The company expects over 100 Ryzen AI Pro commercial platforms in the market by 2025.
7. Client segment revenue grew 29% year-over-year, driven by strong demand for Zen5 processors.
8. Gaming and embedded segments experienced cyclical declines of 69% and 25% respectively.
9. AMD's Versal AI Core adaptive SoCs will power SpaceX's next-generation satellites.
10. The company's long-term thesis remains intact, with potential for significant growth in 5 years.

Takeaways:

1. Trust and customer relationships are crucial for AMD's long-term success in the AI market.
2. AMD's AI GPU business is rapidly expanding, with potential to rival NVIDIA in the future.
3. The company's adaptive technology positions it well for AI at the edge applications.
4. AMD's consistent product improvements and collaborations with major tech companies validate its strategy.
5. Long-term investors may benefit from AMD's potential growth in the AI and computing markets.

Analyzing with edgartools

Let's look at the company using edgartools to pull the data. Getting the company is as simple as using Company("AMD") which gives you basic information about the company.

For AMD, you can see basic information about the industry, which is as expected, semiconductors, and the exchange which is the Nasdaq.

c = Company("AMD")
c

From the company object, you can list the latest 10-Q or 10-K filings. Here we list the last 8 10-Q and 10-K filings for AMD.

c.latest(["10-Q", "10-K"], 8)

If you do not specify the number of filings to return, it will return the latest filing e.g. c.latest("10-Q").

Viewing the latest filing

In edgartools there is the concept called data objects which allow you to upgrade any filing into a data object that contains the data released with that filing. For 10-Q filings that data object is the TenQ which gives you access to the financials released in that filing in that quarter.

filing = c.latest("10-Q")
tenq = filing.obj()
tenq

I think we want to look at the income statement and see how the company is doing generally.

financials = tenq.financials
financials.get_income_statement()

The income statement above is extracted from the XBRL attached to the filing. I have been experimenting with also pulling the financials directly from the HTML document, merging financials across multiple periods, using AI to answer questions and so on.

Conclusion

This was a quick introduction into research companies using edgartools as the means of pulling the company data. I plan a Part 2 where I go in depth into the AMD's financial filings.

If you want to get started with edgartools it's as simple as installing using pip.

pip install edgartools

Star the repo on GitHub!