This article was first featured in Yahoo Finance Tech, a weekly newsletter highlighting our original content on the industry. Get it sent directly to your inbox every Wednesday. Subscribe.
To say Nvidia (NVDA) is on fire is an understatement.
From record revenue and stunning year-over-year growth to share prices that continue to climb and the biggest tech companies on Earth breaking down its door to get access to its latest chips, Nvidia can seemingly do no wrong.
But the chip giant’s rivals, including AMD (AMD) and Intel (INTC), aren’t sitting idly by while it gobbles up the spoils of the AI wave.
They’re busy building out their own AI capabilities in an effort to chip away at Nvidia’s lead. Even Nvidia’s customers, including Google (GOOG, GOOGL), Microsoft (MSFT), and Amazon (AMZN), are building custom AI chips, in order to break free of their dependence on Nvidia’s products.
And while Nvidia is in no immediate danger of losing its grip on the market, Intel’s fall from the world’s most advanced chipmaker proves that there’s no guarantee the company will maintain its lead.
“Things can change pretty dramatically in this industry,” Forrester research director Glenn O’Donnell told Yahoo Finance. “Intel was king of the hill and it stumbled a few years ago and really made some bad mistakes, and now they have to catch up. The same thing could happen to Nvidia.”
Rivals are hunting for their share
Nvidia is the undisputed leader of AI chips, with an estimated 80% to 95% market share, but as its competitors build out their own AI efforts, that’s all but guaranteed to decline.
“I believe on the hardware front, AMD and Intel will be able to field competitive offerings to Nvidia for both training and inference,” Moor Insights and Strategy CEO Patrick Moorhead told Yahoo Finance.
“Both companies will field very interesting offerings available in early 2024 for [large language models (LLMs)]. Even upstarts like Groq have shown advantages on certain models and workflows for LLM inference, today,” Moorhead added.
During Intel’s second quarter earnings report, CEO Pat Gelsinger said the company is seeing surging demand for its AI products and services, including its Gaudi line of AI chips, which will continue into next year.
“Our pipeline of opportunities through 2024 is rapidly increasing and is now over $1 billion and continuing to expand with Gaudi driving the lion’s share,” Gelsinger said during the company’s earnings call.
Intel is also in the midst of developing its next-generation AI platform called Falcon Shores, which it hopes will rival Nvidia’s and AMD’s most powerful systems. Intel initially planned for the chip to hit the market in 2024 and serve as a combined central processing unit (CPU) and graphics processing unit (GPU) platform. Nvidia already offers a combined platform called Grace Hopper.
But Intel has since changed course and Falcon Shores will initially launch with just a GPU. The move also means the chip won’t hit the market until 2025.
AMD is a step ahead of Intel thanks to its MI300 platform. The product line, which includes a combination CPU and GPU offering called the MI300A and a GPU-only chip called the MI300X, could give AMD the kind of capabilities it needs to tackle Nvidia.
The problem? AMD didn’t announce any major customers when it announced the platform in June, according to Reuters. Still, the company and Intel are at least on course to take on Nvidia directly.
“[AMD and Intel are] not going to overcome Nvidia, certainly not anytime in the next few years,” O’Donnell said. “But can they chip away and gain some of that share? Yes, they can. And they already are doing that.”
More companies are building their own chips
Intel and AMD aren’t the only threats to Nvidia’s position at the top of the AI hardware market.
An increasing number of Big Tech firms are building their own custom AI chips. In April, The Information reported that Microsoft was working on an AI chip, and Google, Amazon, and Meta (META) have also announced their own chips.
And with their own AI platforms in hand, those companies won’t necessarily need to buy as many Nvidia chips.
“Hyperscalers [massive cloud firms like Microsoft, Amazon, and Google] are a measurable part of [Nvidia’s] business,” explained Deepwater Asset Management managing partner Gene Munster. “The biggest threat is that these big customers start to try to find their own solutions.”
But Nvidia still has a few aces up its sleeve, chief among them its software capabilities.
“I believe the hardest part of the equation is software, a moat Nvidia has built with CUDA and its libraries and workflows,” Moorhead said.
Nvidia’s CUDA is a programming model that allows developers to take advantage of its chips’ parallel processing capabilities. And it’s that software, coupled with Nvidia’s hardware, that helps give Nvidia a lead over its competitors.
Nvidia’s biggest threats, Intel and AMD, are still years off from being able to put any significant pressure on the graphics chipmaker. But Nvidia will need to continue to innovate if it hopes to stay on top in the long run.
Daniel Howley is the tech editor at Yahoo Finance. Follow him @DanielHowley.
Click here for the latest technology business news, reviews, and useful articles on tech and gadgets
Read the latest financial and business news from Yahoo Finance
