Edge AI stocks compared: Qualcomm (QCOM) vs Arm (ARM) vs Ambarella (AMBA)

Qualcomm, Arm, and Ambarella logos

Key points

  • Three ways to make money from edge AI
  • Qualcomm's fastest-growing chip business
  • Ambarella's bet on vision AI chips

Qualcomm (QCOM), Arm Holdings (ARM), and Ambarella (AMBA) sell into edge AI, which runs models on phones, PCs, cars, and cameras. They make money differently: Qualcomm sells chips and licenses patents, Arm licenses processor designs and collects royalties, and Ambarella sells low-power AI chips for cameras, vehicles, and robots.

The tables compare their businesses and their latest quarterly results. Each company's fiscal year ends in a different month, so their latest quarters end on different dates. Stock prices are September 25 closes.

These results cover each company's entire business. They do not isolate revenue from AI running on devices.

CompanyHow it makes moneyMain edge AI products
Qualcomm (QCOM)Sells chips and licenses its wireless patentsSnapdragon chips for phones, PCs, cars, and connected devices
Arm (ARM)Licenses processor designs and collects a royalty on each chip shipped; introduced its own AGI CPU for data centers in MarchCPU designs and compute subsystems used in phones, PCs, and cars
Ambarella (AMBA)Sells chipsCV-series vision AI chips for security cameras, cars, and robots, plus its first stand-alone AI accelerator, the X7
CompanyQuarter endedRevenueChange from a year earlierGAAP net income (loss)Cash and investmentsSept. 25 close
Qualcomm (QCOM)June 28, 2026$9.95B-4%$2.00B$8.30B$201.97
Arm (ARM)June 30, 2026$1.29B+22%$270M$3.89B$310.32
Ambarella (AMBA)July 31, 2026$108.1M+13%($6.7M)$272.3M$72.68

Figures come from each company's latest quarterly results release. Cash and investments follow each company's own measure, so the balances are not strictly comparable: cash, cash equivalents, and marketable securities for Qualcomm; cash, cash equivalents, and short-term investments for Arm; and cash, cash equivalents, and marketable debt securities for Ambarella.

What is edge AI?

A model that runs on the device itself does not have to send data to a data center and wait for an answer. That can make it faster and keep more data private, but the chip has to do the work within the power and heat limits of a phone, a car, or a camera. All three companies pitch their products on doing more AI work with less power.

Qualcomm: phones are the biggest piece, cars are growing fastest

Qualcomm's chip business brought in $8.50 billion in the quarter ended June 28, according to its results release. Handset chips were the largest part at $5.09 billion, down 20% from a year earlier. Automotive chips rose 61% to $1.59 billion. Chips for connected devices, which Qualcomm reports as IoT, rose 9% to $1.83 billion. Patent licensing added $1.28 billion.

Net income fell 25% to $2.00 billion. The company expects $9.7 billion to $10.5 billion of revenue in the September quarter.

Its newest phone chips put on-device AI at the center of the pitch. Qualcomm says its top Snapdragon chip can run an AI model with more than 30 billion parameters without sending the task to the cloud. Qualcomm is also reaching into data centers, where the company will design custom AI chips for Amazon.

Arm: royalties, licenses, and now its own chip

Arm's revenue rose 22% to a record $1.29 billion in the quarter ended June 30, according to its shareholder letter. Royalties, which Arm collects on each chip its customers ship, rose 22% to $715 million. Arm says newer designs such as Armv9 and its compute subsystems carry higher royalty rates per chip, and that data center royalties more than doubled. License and other revenue rose 23% to $574 million.

In March, Arm introduced the AGI CPU, its first production chip, aimed at data centers, a separate market from edge devices. On the edge side, Arm says AI is moving into PCs, smartphones, vehicles, and robots, and the company points to Nvidia's RTX Spark PC, which uses Arm's compute subsystems.

Operating income was $91 million. Net income was larger, at $270 million, mainly because of $128 million of income from equity investments. Arm held $3.89 billion of cash and short-term investments.

Ambarella: a small company focused on vision AI

Ambarella's revenue rose 13.2% to $108.1 million in the quarter ended July 31, according to its results release. Its net loss narrowed to $6.7 million from $20.0 million a year earlier. The company expects $115 million to $124 million of revenue in the quarter ending October 31.

"Our edge AI revenue reached record levels in Q2, with balanced sequential growth in Auto and IoT markets with very strong growth from our 5nm CV75 and CV72 AI SoCs," said chief executive Fermi Wang. Ambarella also introduced the X7, its first stand-alone AI accelerator, and says more than 50 million of its AI chips are in use. Ambarella held $272.3 million of cash and marketable debt securities.

Frequently asked questions

What is the difference between Qualcomm, Arm, and Ambarella in edge AI?

Qualcomm sells Snapdragon chips for phones, PCs, cars, and connected devices and also licenses its wireless patents. Arm licenses processor designs and collects a royalty on each chip its customers ship, and in March 2026 Arm introduced its own AGI CPU for data centers. Ambarella sells low-power vision AI chips for security cameras, cars, and robots.

How did Qualcomm, Arm, and Ambarella do in their latest quarters?

Qualcomm's revenue fell 4% to $9.95 billion in the quarter ended June 28, 2026, with net income of $2.00 billion. Arm's revenue rose 22% to $1.29 billion in the quarter ended June 30, with net income of $270 million. Ambarella's revenue rose 13% to $108.1 million in the quarter ended July 31, with a net loss of $6.7 million.

Which part of Qualcomm's business is growing fastest?

Automotive. Qualcomm's automotive chip revenue rose 61% to $1.59 billion in the quarter ended June 28, 2026, while handset chip revenue fell 20% to $5.09 billion. Handsets remained its largest chip business.

More on QCOM and ARM

Dennis Singleton
Dennis Singleton

Dennis Singleton was born in Australia and later moved to the United States. He has spent years following the markets, but what keeps his attention is how AI is built. He writes about the companies behind the technology, from semiconductor designers and advanced packaging to photonics, memory, networking, and the hardware powering modern AI. His approach starts with filings, earnings, and industry research, then translates the important details into clear, straightforward analysis without unnecessary hype.