Arm, Broadcom, Nvidia in spotlight as William Blair starts coverage on trio

Arm Holdings (NASDAQ:ARM), Broadcom (NASDAQ:AVGO) and Nvidia (NASDAQ:NVDA) were in the spotlight on Wednesday as research firm William Blair initiated coverage on all three with an Outperform rating.

Shares of all three were little changed in premarket trading.

Arm Holdings

Arm is seen as a “critical vendor” of computing intellectual property with “best-in-class” financials, analyst Sebastien Naji wrote in an investor note, adding that the company generates revenue from the more than 29B chips sold in the mobile, automotive, IoT, and data center markets.

“Arm’s royalty/licensing revenue model drives best-in-class profitability—R&D is the largest expense, 35% of total revenue in fiscal 2024,” Naji wrote. “With expanding royalty rates helping drive better operating leverage (long-term target of 60% non-GAAP operating margin), we see room for sustained EPS and free cash flow growth.”

The British chip design firm, which is still majority owned by Japanese tech conglomerate SoftBank (OTCPK:SFTBY), released the newest version of its processor design, v9. The newest design has “substantial improvements” over the prior generations — including encryption and vector processing — which allowed Arm to “double” its take-rate compared to the v8 design and should boost royalty revenue growth over the next three to four years, Naji said.

The Rene Haas-led company should also benefit from the introduction of other subsystems, including CSS for mobile and Neoverse for data center, both of which have helped drive new licensing activity and should increase Arm’s take-rate again as shipments rise in 2025.

Arm is also starting to see traction in the data center market, which Intel (INTC) has historically dominated. But as hyperscalers like Amazon (AMZN), Google (GOOG) (GOOGL), Microsoft (MSFT) and Meta (META) build their own chips for their artificial needs and Nvidia (NVDA) dominates the AI accelerator market, the opportunity for Arm to keep benefiting is there, Naji said.

“As these chips are manufactured at scale and built into ever-growing data centers, we see substantial room for Arm’s data center business to become a more significant growth driver and higher portion [of] the royalty revenue base (we estimate 15% mix in fiscal 2025),” Naji explained.

Broadcom

Broadcom has become a dominant player in the AI space thanks to its custom chips business (which develops Google’s tensor processing units and Meta’s MTIA accelerators) and its Ethernet networking business.

“We see room for continued steady growth going into fiscal 2025 and 2026 driven by increasing custom chip demand, improved software monetization, recovery in non-AI semis, and accelerating growth of Ethernet AI network fabrics built on top of Broadcom’s networking solutions—Ethernet is just starting to displace InfiniBand as the solution of choice for AI networks,” Naji said.

Broadcom’s recent acquisition of VMware should also help boost the company’s software revenue, as it focuses on raising licensing costs and keeping its highest return on investment customers.

“AVGO shares trade at a P/E multiple of 26 times and an EV/FCF multiple of 24 times our calendar 2025 estimates, below the peer group median multiples,” Naji wrote. “Given the multiple growth tailwinds and best-in-class financial model, we see room for upside to both multiples and earnings.”

Nvidia

Nvidia is already the leader in the AI space and, in particular, high-performance computing, as its GPUs have become synonymous with the generative AI boom. However, the company’s technological prowess and continued innovation should allow it to stay ahead of competitors for some time, Naji said.

“Nvidia’s technical differentiation extends beyond building state-of-the-art processors to include fully integrated systems,” Naji wrote, noting that it has a “deep” software ecosystem with more than 5M active developers for CUDA and competencies networking (via its Mellanox and Cumulus acquisitions), systems engineering, and supply chain management.

“This vertical approach to computing is packaged into Nvidia’s DGX product, which has greatly expanded its [total addressable market] beyond the traditional GPU market (roughly $100 billion TAM) to the broader semiconductor market ($800 billion) and cloud services markets ($1.6 trillion),” Naji said.