NEW YORK, May 27, 2026, 11:02 EDT
- Options pricing is signaling Broadcom shares could move about 7.5% when the company reports next week.
- Investors are looking at rapid AI-chip growth but also asking if custom chips can deliver Nvidia-like margins.
- Broadcom slipped in early U.S. trading but stayed above the $2 trillion mark on market value.
Broadcom (AVGO) options imply the stock could move around 7.5% in either direction following its quarterly earnings, due after the close on June 3, Bloomberg data shows. That sets up next week’s results as a key moment for one of the market’s top AI names.
Broadcom is shifting from a reputation as a broad chip and software maker to one focused on custom AI silicon. Its market cap stood at about $2.04 trillion Wednesday. Shares were off about 0.5% to $419.70 in the morning. Nvidia dropped 2.3% and Marvell fell 2.8%, putting pressure on AI chip stocks.
Broadcom is set to release its fiscal Q2 2026 numbers and outlook after the bell on June 3, the company said. Management will hold a call with analysts at 5 p.m. ET.
The Broadcom stock story isn’t just about demand now. The new debate is if Broadcom’s custom accelerators, designed for individual clients instead of sold as standard chips, can grow quickly enough without cutting into margins.
Broadcom reported in March that first-quarter revenue climbed 29% to $19.31 billion. The company guided for second-quarter revenue of around $22 billion, which would be a 47% jump from the same period last year. CEO Hock Tan said AI revenue growth was “accelerating” and forecast $10.7 billion in AI semiconductor revenue for the second quarter. Broadcom Inc.
Volatility is back in focus. According to Investing.com, which cited Bloomberg options data, Broadcom stock has swung more than options markets implied in half of its last eight earnings reports. The outsized moves have gone both up and down: shares jumped 31.8% in December 2024 and 26.2% in June 2024, but they also fell 13.4% in September 2024 and dropped 8.4% in March 2025.
Broadcom is pitching itself to customers looking for something other than Nvidia’s general-purpose GPUs, which have dominated AI training and inference. Back in March, Reuters said that Big Tech’s AI infrastructure spending could go over $600 billion this year, boosting demand for chips, servers, storage, and networking gear.
Marvell is getting drawn into the custom-chip battle. The Register reported Wednesday that IP firms like Broadcom and Marvell have built big chunks of custom AI silicon for hyperscale data centers. A lot of this activity hasn’t been public until now.
Broadcom (AVGO) is teaming up with FuriosaAI on a third-generation AI accelerator for inference, the companies said Wednesday. The chip will be built with a 2-nanometer compute die, HBM4/4E memory, plus Ethernet and PCIe tech from Broadcom. Sampling should begin in the first half of 2028.
Broadcom’s Semiconductor Solutions Group president Charlie Kawwas said inference now depends more on “data reuse and communication efficiency” across servers and racks. Furiosa co-founder and CEO June Paik called the deal a step past the chip level, saying it brings the product into the “token factory era.” Stock Titan
But the main risk is margin, not just sales. Barchart columnist Mikhail Fedorov wrote that custom AI chips are a different business than Nvidia’s GPU model, since the end chip is usually owned by the customer, with Broadcom handling architecture, IP, and supply chain work. In this arrangement, the customer is focused on cutting infrastructure costs, which he said could keep a lid on Broadcom’s profits as custom silicon accounts for a bigger part of revenue.
Broadcom’s bigger software business, boosted by VMware, acts as a buffer few chip-focused rivals have. For the first quarter, infrastructure software brought in $6.80 billion, or 35% of Broadcom’s revenue, with semiconductor solutions at $12.52 billion, about 65%. CFO Kirsten Spears said adjusted EBITDA reached 68% of sales and kept the forecast for that margin in the second quarter.
June 3 is now set up as the key test for whether investors see real earnings leverage or just more AI hype. The stock doesn’t need a flawless report, but at a $2 trillion valuation and options ready for a big swing, Wall Street wants specifics on AI revenue, margins, supply, and customer mix.
