Semiconductor companies NVIDIA (NASDAQ: NVDA) The company dominates the market for artificial intelligence (AI) chips — in fact, analysts estimate that the company’s graphics processing units account for up to 95% of AI chip sales — but history is full of lessons about the downfall of major brands that failed to sustain technological leadership.
IBM Once a computing company prized for its mainframe and storage solutions, it failed to innovate on cloud computing and PCs. Cisco At the time the spread of the Internet increased demand for networking products, the company was once the world’s most valuable company, but due to its failure to innovate, it is no longer ranked in the top 50.
These are just two examples in a long list that includes AOL. blackberriesBlockbuster, Kodakand MySpace. To be clear, Nvidia is in no danger of joining that list anytime soon, but history has shown that shareholders would be foolish to completely dismiss the idea. apple (Nasdaq: AAPL) and Amazon (Nasdaq: AMZN) Let’s be clear that Nvidia is not immune from competition.
Here are the important details.
Apple issues warning to Nvidia shareholders
Earlier this year, Apple artificial intelligence (AI) capabilities for iOS and macOS devices called Apple Intelligence, which includes features like text drafting and correction, image generation, and notification summaries, as well as a more capable version of the personal assistant Siri. Apple Intelligence will be released in late October.
In July, Apple published a technical paper stating, “Apple Intelligence consists of multiple high-performance generative models that are fast, efficient, specialized for your everyday tasks, and instantly adapt to your current activity. The paper also details how two of these models were trained: one that runs on-device for simple tasks, and one that runs in private data centers for more advanced tasks.”
The key is that Apple doesn’t use Nvidia. Graphics Processing Unit (GPUs) to develop large language models. The company uses custom silicon called tensor processing units (TPUs) to alphabetGoogle and BroadcomNotably, Broadcom also helps other companies design custom AI chips, including Meta PlatformAccording to , OpenAI and ByteDance, JPMorgan Chase.
The bottom line is this: Apple’s decision to use TPUs rather than GPUs to train its AI models is a warning that viable alternatives to Nvidia exist. The relative importance of this decision remains to be seen. If Apple Intelligence has flaws, it could serve as a cautionary tale of what happens when companies stray from industry standards. But if Apple Intelligence is well-received by consumers, Apple’s decision to use TPUs could encourage other companies to do the same.
Amazon warns NVIDIA shareholders
Alphabet’s Google isn’t the only public cloud developing custom AI silicon: Amazon Web Services (AWS) is doing the same. Its custom chips — Trainium for AI training and Inferentia for AI inference — are designed not to outperform Nvidia GPUs on performance alone, but to offer customers a more cost-effective alternative.
Alphabet and Amazon have close partnerships with Nvidia. Both offer compute instances with Nvidia GPUs, and that’s not going to change anytime soon. Nvidia GPUs are the gold standard for data center accelerators, so it would make no sense not to offer that option. However, both companies see the opportunity to develop their own chips.
Amazon CEO Andy Jassy recently said:
“We’ve heard loud and clear from customers that they would welcome better price performance, which is why we’ve invested in our own custom silicon, Trainium for training and Inferentia for inference. And the second version of these chips, Trainium, which will be released later this year, offers very good price performance. We’re seeing strong demand for these chips.”
The bottom line is this: While some companies will be happy to pay a premium for Nvidia GPUs, AWS’s demand for custom silicon suggests that others are happy to use slower chips to cut costs. In other words, AWS (like Apple) is warning investors that Nvidia is almost certain to lose market share in the future, potentially squeezing the company’s profit margins with competition from cheaper alternatives.
Should I invest $1,000 in Nvidia right now?
Before you buy Nvidia stock, consider the following:
of Motley Fool Stock Advisor The analyst team Top 10 Stocks Here are the stocks investors should buy right now… and Nvidia wasn’t among them. The 10 stocks selected have the potential to generate big gains over the next few years.
Things to consider NVIDIA This list was created on April 15, 2005…If you invested $1,000 at the time of recommendation, That works out to $630,099.!*
Stock Advisor With portfolio construction guidance, regular updates from our analysts, and two new stock picks every month, we provide investors with an easy-to-follow blueprint for success. Stock Advisor The service is More than 4 times First S&P 500 recovery since 2002*.
*Stock Advisor returns as of September 9, 2024
John Mackey, former CEO of Amazon subsidiary Whole Foods Market, is a member of The Motley Fool’s board of directors. Alphabet executive Suzanne Frey is a member of The Motley Fool’s board of directors. JPMorgan Chase is an advertising partner of The Ascent, a Motley Fool subsidiary. Randi Zuckerberg, former director of market development and public relations at Facebook and sister of Meta Platforms CEO Mark Zuckerberg, is a member of The Motley Fool’s board of directors. Trevor Genevin The Motley Fool has invested in and recommends Alphabet, Amazon, Apple, JPMorgan Chase, Meta Platforms, and Nvidia. The Motley Fool recommends BlackBerry, Broadcom, and International Business Machines. The Motley Fool recommends Disclosure Policy.
Nvidia stock investors receive stern warnings from Apple and Amazon Originally published on The Motley Fool
