Established in 1993 by Tom and David Gardner, The Motley Fool is dedicated to making the world smarter, happier, and richer. With our website, podcasts, books, newspaper column, radio show, and premium investing services, we strive to help millions of individuals achieve financial freedom. In addition to our own AI chips, Intel has the potential to manufacture advanced AI chips for third parties. While reading this free article that presents varying opinions, it's important to note that they may differ from The Motley Fool's Premium Investing Services. To gain immediate access to our top analyst recommendations, in-depth research, investing resources, and more, become a member of Motley Fool today. Learn More Training sophisticated AI models like the one powering OpenAI's ChatGPT not only requires substantial processing power but also the right kind of processing power. Nvidia's graphics processing units (GPUs) are well suited for this task. With thousands of GPUs linked together in their data centers, Nvidia can efficiently process vast datasets and bring large language models to life. Although Nvidia's GPUs have become the industry standard for AI acceleration, Intel is approaching the market from multiple angles. The company offers its own data center GPUs, including the recently launched Flex and Max series products. Intel's Max GPUs will be utilized in the Aurora supercomputer at Argonne National Laboratory, scheduled to go online this year. Additionally, Intel's acquisition of Habana Labs in 2019 has positioned its AI accelerator business at the forefront. The Gaudi2 AI chip, which has demonstrated impressive performance in recent AI benchmarks, is gaining rapid traction during the AI boom. During Intel's second-quarter earnings call, CEO Pat Gelsinger shared that the AI accelerator pipeline has surpassed $1 billion and is growing exponentially. The primary driver of this soaring demand is the Gaudi2 chip. Intel is currently shipping Gaudi2 at scale, and Amazon Web Services offers Gaudi instances. Although Nvidia's latest H100 data center GPU outperforms Gaudi2 in raw performance, a specialized chip like Gaudi2 can effectively compete in terms of total cost of ownership and efficiency, as it is specifically designed for AI training workloads. Intel has secured commitments for its Gaudi product line from major players in the AI industry, including Hugging Face and Stability AI. Looking ahead, the company plans to introduce new products to further expand its AI accelerator business.

The next-generation Gaudi3 chips are set to launch in 2024, followed by the Falcon Shores GPU in 2025 and Falcon Shores 2 in 2026. Intel can also capitalize on the growing demand for AI accelerators through its foundry business. Major cloud companies like Alphabet and Amazon design their own custom AI chips, and while Intel has not provided detailed information, a portion of its foundry pipeline for the upcoming 18A manufacturing process is allocated to customers designing their own AI accelerators. With the expectation of reclaiming its manufacturing lead with 18A, Intel becomes an attractive option for customers seeking the most advanced manufacturing process. While Intel's core businesses are still focused on PC and server CPUs, both segments have encountered challenges. However, there are signs that the decline in demand is stabilizing. After a series of losses, Intel managed to turn a profit in the second quarter, indicating that the worst may be behind them. AI accelerators have the potential to become a significant business for Intel in the coming years as companies seek alternatives to Nvidia's costly GPUs. Accelerators offer a multibillion-dollar opportunity, and manufacturing accelerators for third parties further enhances this potential. With Intel's stock currently at discounted levels, it presents an appealing opportunity to invest in the AI revolution. It is important to note that Suzanne Frey, an executive at Alphabet, is a member of The Motley Fool's board of directors, while John Mackey, the former CEO of Whole Foods Market, an Amazon subsidiary, also serves on The Motley Fool's board of directors. Timothy Green holds positions in Intel. The Motley Fool has positions in and recommends Alphabet, Amazon. com, and Nvidia. Furthermore, The Motley Fool recommends Intel and provides the following options: long January 2023 $57. 50 calls on Intel and long January 2025 $45 calls on Intel. The company adheres to a disclosure policy. The calculation is based on the average return of all stock recommendations since the inception of the Stock Advisor service in February 2002. Returns are accurate as of July 29, 2023. Discounted offers are exclusively available to new members, while the list price for Stock Advisor is $199 per year. The volatility profiles are based on three-year calculations of the standard deviation of service investment returns.