The development of cutting-edge artificial intelligence (AI) models would not be possible without the semiconductor industry.
In 2016, NVIDIA (NVDA 2.48%) CEO Jensen Huang personally delivered OpenAI’s first artificial intelligence (AI) supercomputer, which, in retrospect, was a key moment in history because the startup would go on to develop one of the AI industry’s most advanced applications: ChatGPT.
ChatGPT and many of its competitors can already generate text, images, videos and computer code in response to commands, but Nvidia and its peers are designing more powerful and energy-efficient hardware, paving the way for the development of even better AI applications with expanded capabilities.
Here are five semiconductor stocks that investors might want to buy amid the AI revolution.
Image source: Getty Images.
1. NVIDIA
Nvidia is best known for designing the most powerful datacenter graphic processing units (GPUs) for AI development, but the Nvidia AI Enterprise platform goes beyond just chips: it’s an entire cloud-based operating system for AI developers.
The platform provides developers with a library of ready-made large language models (LLMs) that they can use to build AI applications, saving them significant time and money. It also features CUDA, a software layer for GPUs, which allows developers to optimize the chip to build applications faster. CUDA is not available on other chips, so data center operators must stick with Nvidia hardware or risk angering software-savvy customers.
Nvidia’s H100 GPU set the AI industry benchmark, but the company is preparing to ship a new generation of chips based on its Blackwell architecture. For example, the upcoming GB200 GPU can run AI inference five times faster than the H100, lowering costs for developers who often pay by the minute for computing power.
Nvidia’s stock has tripled in the past year alone, and it probably won’t stop rising. Wall Street expects the company to bring in more than $120 billion in revenue in its current fiscal year 2025 (ending Jan. 30, 2025), nearly double what it made in fiscal year 2024. Most of that revenue will come from its AI-focused data center division.
2. Advanced Micro Devices (AMD)
Nvidia is struggling to meet demand for its GPUs, with some major customers Advanced Micro Devices (AMD -0.89%) Instead, AMD’s MI300 GPU has the backing of data center giants such as: Microsoft, Oracleand Meta Platform,We found that it outperforms H100 in terms of inference performance and cost.
AMD’s data center revenue grew 80% to $2.3 billion in the most recent first quarter of 2024, which ended March 31. The company expects to achieve $4 billion in sales from GPUs alone this year, up from a January forecast of $3.5 billion.
But AMD is also taking a leadership position in another key area of AI: personal computing: The company says millions of PCs with its Ryzen AI chips have shipped to date from major manufacturers. Dell, home pageAsus, etc. AMD has 90% market share in this field.
The company is working closely with Microsoft to develop new Ryzen chips that can support the growing number of AI features built into the Windows operating system.
AMD shares are up 51% over the past year but remain 18% below their all-time high. With the company’s AI revenue growing in both the data center and PC segments, now could be a good time to buy.
3. Axcelis Technologies
Unlike Nvidia and AMD, Axcelis Technologies (ACLS 5.17%) The company doesn’t make chips. It makes ion implantation equipment, which is essential to making processors (CPUs), memory chips, and storage chips. AI applications require high capabilities in all three, which represents a big opportunity for the company.
Most of Nvidia’s datacenter GPUs come with built-in memory, and advanced models such as the Blackwell GB200 also have built-in CPUs for greater efficiency.
What’s more, AI-enabled PCs and smartphones require more processing power and up to twice as much memory as their predecessors. As a result, Axcelis CEO Russell Low says AI will drive a massive expansion of manufacturing capacity across the semiconductor industry, which should boost sales of his company’s equipment.
Additionally, the manufacturing of power devices – the chips that control the flow of power in high-current workloads – requires a huge number of implants. The demand for AI data centers is growing rapidly, driving up the demand for energy generation, distribution and cooling, which in turn is driving up the demand for Axcelis’ equipment.
Despite soaring 835% over the past five years, the stock is trading at a price-to-earnings (P/E) ratio of just 18.7, a 49% discount to its P/E of 36.9. iShares Semiconductor ETFIn other words, Axcelis’ stock price would have to nearly double to trade at the same level as its semiconductor peers, which represents an opportunity for investors.
4. Broadcom
Broadcom (AVGO -0.72%) is a multifaceted AI company that makes hardware and software networking solutions for data centers, including the switches that control the speed at which data moves between servers and devices. These switches are a key component in clustering thousands of GPUs to process massive amounts of data in AI development.
In the recent fiscal second quarter (ended May 5), Broadcom’s sales of its Tomahawk 5 and Jericho 3 switches doubled year over year.
Many of the company’s subsidiaries are also using AI, including Symantec, which it acquired for $10.7 billion in 2019. alphabetIt has partnered with ‘s Vertex AI platform to bring AI to its cybersecurity software, and Broadcom acquired cloud software provider VMware for $69 billion in 2023, which allows developers to create virtual machines, meaning multiple users can connect to a single server and use all of its capacity, important in an era of scarce AI infrastructure.
Broadcom announced that its company-wide AI revenue reached $3.1 billion in the second quarter, up 280% year over year. For the full fiscal year 2024, the company expects to achieve a record $51 billion in total revenue, of which $11 billion will come from AI.
5. Micron Technology
As mentioned earlier, AI-enabled PCs and smartphones require significantly more memory capacity than non-AI devices because AI applications require huge amounts of data to function, and memory chips act as the brains that temporarily store that information in a ready state. Micron Technology (MU 0.34%) is a leading provider of memory and storage chips and is experiencing a wave of demand as a result of the AI revolution.
All Tier 1 manufacturers of Android-based AI smartphones ( Samsung) use Micron’s LPDDR5X memory chips, with capacities ranging from 12 gigabytes to 16 gigabytes, which is 50% to 100% more memory than previous flagship smartphones required. Additionally, Microsoft’s new Copilot+ AI PC has a minimum memory requirement of 16 gigabytes, double the minimum required for the previous Surface lineup. This trend means increased revenue for Micron.
On the data center side, Micron’s HBM3e (high-bandwidth memory) solutions are used in Nvidia’s new H200 GPUs, which can run AI inference nearly twice as fast as the H100 and consume half the power, resulting in significant cost savings for data center operators.
Micron said HBM3e contributed $100 million to its revenue in the third quarter of fiscal 2024, which ended May 30. By the end of fiscal 2024, the company expects to generate hundreds of millions of dollars worth of sales. Billions In fiscal year 2025, it will be dollars.
Micron has already sold out of HBM3e memory for the next two years, which gives the company incredible pricing power and should significantly improve its profitability.
