Broadcom: A Potential Millionaire-Maker in the AI Chip Race

Is Broadcom a Millionaire-Maker Stock? | The Motley Fool

If you invested a mere $10,000 in Broadcom (AVGO 2.20%), then called Avago Technologies, shortly after the stock went public in 2009, and never sold any shares, you'd be pretty close to having a million-dollar investment in just 15 years' time. While Broadcom may not put up the same type of returns over the next 15 years, the stock still has the potential to be a millionaire maker. While rival Nvidia has been the biggest beneficiary of the early artificial intelligence (AI) infrastructure buildout, Broadcom is well positioned to become the next big AI winner. That's not to say, however, that the company isn't already doing well in this area. For example, the company has continued to push up its forecast for AI revenue throughout the year. Last fall, it expected to generate around $7.5 billion in AI revenue for 2024. It later bumped that up to $10 billion and then $11 billion. After its most-recent quarter, it increased its expectations for AI revenue to $12 billion. The company participates in the buildout of AI data-center infrastructure in two main ways: networking components and custom chips. On the networking side, the company designs a number of components that go into the building of graphics processing units (GPUs) clusters, which are a group of interconnected computers that work together as a single system that has GPUs on each node (an individual server within the cluster). It is these supercomputer GPU clusters that are being used to train AI large language models (LLMs). While GPUs get all the press given that the number of them going into these clusters continues to grow, they are not the only components needed. On Broadcom's end, it supplies the switches and network interface cards (NICs) that are needed to build these clusters. Switches allow two or more devices to communicate directly with each other, help manage the flow of data, and avoid network congestion. NICs, meanwhile, are needed to connect computers to a network to be able to communicate with other computers on the network. Clusters continue to grow exponentially in size with more and more GPUs. Broadcom sees this growth creating a distributed-computing challenge for which Ethernet switches will be at the heart of handling AI workloads and transferring data between GPUs. Broadcom is not without competition. Nvidia offers a competing switching technology to Ethernet called InfiniBand, which it got as part of its acquisition of Mellanox, while Intel has developed a technology called Omni-Path. For its part, Broadcom said it thinks that all hyperscalers will be using Ethernet by the first half of 2025. Perhaps the even bigger opportunity moving forward for Broadcom, though, is through the design of custom AI chips (application-specific integrated circuits, or ASICs) for customers. Its first customer in this area was Alphabet, for which it helped create the company's tensor-processing unit for AI workloads. It has added two additional customers this year, which are believed to be Meta Platforms and China's ByteDance, owner of TikTok. There are also reports that OpenAI has been in discussions with Broadcom to build its own custom AI chip. With Broadcom's ASICs, customers are able to customize the chips to meet their specific computing and power needs. This is a huge and growing market, and Broadcom is a leader in the ASIC space. With AI capital expenditure (capex) budgets increasing, Broadcom is very well positioned to capture what could be the next leg of the AI infrastructure buildout: custom AI chips. Broadcom appears to be in the early innings of the custom AI chip revolution. This has the opportunity to be a huge market, and the company is the leader in the ASIC space. From a valuation perspective, the company now trades at a forward price-to-earnings (P/E) ratio of about 26 based on next year's analyst estimates. Given the growth opportunities in front of the company, that's not an expensive valuation. If Broadcom can capitalize on its custom AI chips and advanced networking opportunities, the stock has the makings to be the next big AI winner and a potential millionaire maker.

3 Reasons to Buy Broadcom Stock Like There's No Tomorrow | The Motley Fool

Broadcom's custom AI chips are in high demand, and it looks like the market is just getting started. Investing in artificial intelligence may feel a little wild right now. So much is changing, and many tech companies that weren't formerly in the AI space have pivoted to the market (or at least tried to). But there's one company that is seemingly making the right moves in AI right now, benefiting from a unique niche in the market, and has the potential to continue tapping into AI in the years ahead. Here's why Broadcom (AVGO 2.20%) could be a great addition to your portfolio. Everyone who's been following along at home knows that Nvidia is the leader in AI graphics processing chips. However, within the broader AI chip market, there are niches, including application-specific integrated circuits (ASICs), that are up for the taking. This is where Broadcom has focused its attention lately. The company already makes the leading ASICs used by companies needing general-purpose AI chips (think big tech companies like Alphabet and Meta), and is attracting a growing list of customers in an expanding market. Because of Broadcom's early moves in this space, J.P. Morgan estimates the company's custom AI chips have a total addressable market of $150 billion over the next four to five years. Broadcom isn't just waiting for the large total addressable market to help boost revenue one day; it's already benefiting from it. Sales of the company's AI-focused chips more than tripled in the second quarter, reaching $3.1 billion. Not only is that phenomenal year-over-year growth, but considering that total revenue in the quarter was $13.1 billion, Broadcom's AI sales accounted for an impressive 24% of the company's sales in the quarter. Chip customers keep returning to Broadcom, and CEO Hock Tan said on the earnings call that "customers continue to scale up and scale out their AI clusters," and that custom AI accelerator demand continues to grow. The increasing demand for its AI products prompted Broadcom's management to raise its artificial intelligence revenue estimates for 2024 from the previous $11 billion to $12 billion. Admittedly, Broadcom's stock isn't cheap. But compared to Nvidia and even AMD, it's slightly cheaper. Broadcom's current forward price-to-earnings ratio is 27, while AMD's is 28 and Nvidia's P/E ratio is a whopping 42. For investors looking for a better deal than some other AI chip plays, Broadcom is about as good as you're going to get. Broadcom's share price dipped after the company reported its third-quarter results because fourth-quarter guidance was a little lower than analysts' estimates. With shares down about 6% over the past three months, now could be a good time to buy the stock. Just keep in mind that AI stocks can be volatile, so expect some price swings along the way.

Better Artificial Intelligence Stock: Nvidia vs. Broadcom | The Motley Fool

Over the past year, businesses raced to implement artificial intelligence (AI) technologies, and that boosted sales for semiconductor chip makers Nvidia (NVDA -1.59%) and Broadcom (AVGO 2.20%). Both sell components required to implement AI in data centers designed for cloud computing, the ideal place for AI to satisfy its need for massive computing power. As a result, Broadcom and Nvidia are positioned to benefit for years as the AI industry expands from 2023's global market size of $136 billion to an estimated $827 billion by 2030. This is one reason owning shares in both is a great way to invest in the secular trend of AI. However, if you had to choose just one of these semiconductor giants to invest in, which would make the better AI investment? Here's a breakdown of each to arrive at an answer. Broadcom tackles the AI market with an array of products servicing many tech industries, including data centers, wireless networks, and industrial applications. The company's AI strategy involves a combination of hardware and software to help customers meet their needs for AI infrastructure. On the hardware side, Broadcom's products include AI accelerators, an essential component to speed up AI computer processing. Its business building custom AI accelerators grew year over year by more than 3 times in its fiscal third quarter, ended Aug. 4. Thanks to demand for its AI-related products, the company's semiconductor division saw fiscal Q3 sales grow to $7.3 billion, up from $6.9 billion in the previous year. On the software side, Broadcom acquired VMware toward the end of 2023, which is why its software division's Q3 revenue zoomed up to $5.8 billion from $1.9 billion in the prior year. The addition of VMware enabled total Q3 revenue to grow 47% year over year to $13.1 billion. Broadcom is using the VMware Cloud Foundation (VCF) platform to target businesses wanting a private cloud, or a hybrid of private and public clouds, for their AI systems. The strategy is succeeding. VCF represented 80% of VMware's products booked in Q3. Broadcom exited its fiscal third quarter with respectable financials, such as free cash flow (FCF) of $4.8 billion. Its balance sheet included total assets of $168 billion, with $10 billion of that in cash and equivalents. Total liabilities were $102.3 billion, with $70 billion of that in debt. Nvidia's AI success centers around its graphics processing unit (GPU), which is the heart of an AI system's ability to execute complex computations. The company also provides a comprehensive suite of software and other services, such as a robotics platform, to help businesses harness AI. The semiconductor giant pioneered the GPU in 1999, steadily improving it over the years. Now, Nvidia's chips are the preferred choice for the AI computing industry. The company continues to evolve its processors, with its new Blackwell GPUs ramping up production this year. According to Nvidia, Blackwell is the world's largest, most powerful GPU, housing over 200 billion transistors. Its GPU leadership in AI enabled Nvidia to hit a record $30 billion in revenue during its fiscal second quarter, ended July 28, a 122% increase from a year ago. The company expects sales to keep growing, forecasting fiscal Q3 revenue to reach $32.5 billion, up from the prior year's $18.1 billion. Nvidia's AI strategy includes extending its GPU strength into various industries, such as automotive, to enable self-driving vehicles. In addition, its CEO, Jensen Huang, believes the cloud computing market, worth nearly $600 billion in 2023, will transition existing infrastructure to support AI systems. This would be a massive market opportunity for chipmakers such as Nvidia. Adding to the company's investment appeal, Nvidia's financials are excellent. Its fiscal Q2 FCF was $13.5 billion. Its Q2 balance sheet included $85.2 billion in total assets, with $34.8 billion of that in cash, cash equivalents, and marketable securities. Total liabilities were $27.1 billion, with $8.5 billion in debt. Because the AI market is large and growing, both Broadcom and Nvidia should continue to see rising sales. But between the two, Nvidia experienced the largest revenue growth this past year as AI took off, illustrating the customer demand for its products. Data by YCharts. Another factor to consider with each stock is their price-to-earnings ratio (P/E ratio), a widely used metric to assess valuation. Broadcom's P/E multiple of 141 is far higher than the 54 for Nvidia shares at the time of this writing, indicating Nvidia is the better value. Currently, both Broadcom and Nvidia shares are below their 52-week highs. But Nvidia's market leadership in AI chips, strong financials, and superior valuation compared to Broadcom make it the better long-term investment between these two semiconductor titans.