Marketdash

RBC Capital Says the AI Semiconductor Rally Still Has Legs Despite Bubble Fears

MarketDash Editorial Team
7 hours ago
While investors fret about an AI bubble, RBC Capital Markets argues the semiconductor sector has more room to run. Analyst Srini Pajjuri expects hyperscale spending to stay strong for 18-24 months, with Nvidia, Micron, and chip equipment makers positioned to benefit most from continued AI infrastructure buildout.

Get Applied Materials Alerts

Weekly insights + SMS alerts

The AI bubble debate is heating up, but RBC Capital Markets isn't backing away from semiconductors just yet. In fact, the firm sees plenty of runway left in the sector, provided you know where to look.

Analyst Srini Pajjuri argues that despite growing concerns about overheated AI valuations, the semiconductor outlook remains intact. The key driver? Hyperscale capital spending should stay strong for the next 18 to 24 months as tech giants battle for AI supremacy. The competitive dynamics alone make it hard for any single player to pull back without losing ground.

Here's what matters: AI monetization keeps improving, and the balance sheets of the largest hyperscalers don't look stressed, even with infrastructure bottlenecks and project delays popping up. Pajjuri expects any eventual spending slowdown to happen gradually rather than falling off a cliff. His advice? Keep balanced exposure to AI-focused semiconductor names, but be selective.

The GPU Heavyweights

Graphics processing units remain the backbone of the AI infrastructure buildout, and Nvidia Corp. (NVDA) sits firmly at the center of it all.

Pajjuri sees limited threats to Nvidia's full-stack dominance, even as custom Application-Specific Integrated Circuits and rival GPUs make progress. Hyperscaler spending should support Nvidia's order visibility going forward, and the analyst believes the valuation already bakes in some slowdown risk. Translation: there's still upside if things go right.

Advanced Micro Devices, Inc. (AMD) gets credit as a credible secondary GPU supplier, especially after landing a deal with OpenAI. But there's a catch. AMD's valuation already reflects near-term production ramps, which means upside may be limited until later product cycles kick in.

As for Broadcom Inc. (AVGO), Pajjuri acknowledges momentum in custom accelerators but questions whether certain customer programs will last. He also flagged valuation concerns and margin pressures as potential headwinds.

Memory Markets Get a Structural Boost

High-bandwidth memory is turning into a transformative force that could reduce the notorious boom-bust cycles plaguing the memory market. Pajjuri expects HBM to represent roughly a quarter of DRAM industry revenue and grow at over 40% annually through 2028.

Tight supply should keep demand running ahead of supply into 2027, which sets up Micron Technology Inc. (MU) particularly well. The analyst sees persistent DRAM tightness and meaningful earnings upside for Micron as the HBM story plays out.

There's another wrinkle here: GenAI workloads are becoming increasingly memory-intensive, which supports HBM demand. The flip side? Higher memory prices could create headwinds for PCs and smartphones, which are more price-sensitive markets.

Get Applied Materials Alerts

Weekly insights + SMS (optional)

Chip Equipment Makers Stay Busy

Wafer fab equipment spending should maintain an upward trajectory over the next two years, even if China-related spending softens. Pajjuri expects memory-related WFE spending to outpace logic spending, driven by extremely tight supply conditions in memory markets.

The winners here include Lam Research Corp. (LRCX), Applied Materials Inc. (AMAT), and ASML Holding NV (ASML). These names are positioned to benefit as memory manufacturers race to add capacity and meet surging HBM demand.

The Supporting Cast

Beyond the marquee names, Pajjuri highlighted several other beneficiaries of the AI infrastructure buildout.

Marvell Technology Inc. (MRVL) stands out among networking and connectivity plays, while Astera Labs benefits from AI system interconnect trends as data centers get more complex.

Arm Holdings Plc (ARM) gets a nod for its expanding data center footprint and pricing power, which should support long-term growth even as smartphone demand faces memory-related pressures.

The Bottom Line

Pajjuri's overarching thesis is that AI has split the semiconductor sector into clear winners and losers. Selectivity matters now more than ever, especially with valuations sitting above historical averages. Any meaningful slowdown in hyperscale spending could hit stock prices hard, so understanding which companies have durable competitive positions becomes critical.

The message isn't that everything will work out perfectly. It's that the structural drivers remain intact, the spending environment should stay supportive for at least another year or two, and the right names still offer meaningful upside if you're willing to do the work to separate the haves from the have-nots.

NVDA Price Action: Nvidia shares were up 2.96% at $188.57 at the time of publication on Thursday, according to market data.

RBC Capital Says the AI Semiconductor Rally Still Has Legs Despite Bubble Fears

MarketDash Editorial Team
7 hours ago
While investors fret about an AI bubble, RBC Capital Markets argues the semiconductor sector has more room to run. Analyst Srini Pajjuri expects hyperscale spending to stay strong for 18-24 months, with Nvidia, Micron, and chip equipment makers positioned to benefit most from continued AI infrastructure buildout.

Get Applied Materials Alerts

Weekly insights + SMS alerts

The AI bubble debate is heating up, but RBC Capital Markets isn't backing away from semiconductors just yet. In fact, the firm sees plenty of runway left in the sector, provided you know where to look.

Analyst Srini Pajjuri argues that despite growing concerns about overheated AI valuations, the semiconductor outlook remains intact. The key driver? Hyperscale capital spending should stay strong for the next 18 to 24 months as tech giants battle for AI supremacy. The competitive dynamics alone make it hard for any single player to pull back without losing ground.

Here's what matters: AI monetization keeps improving, and the balance sheets of the largest hyperscalers don't look stressed, even with infrastructure bottlenecks and project delays popping up. Pajjuri expects any eventual spending slowdown to happen gradually rather than falling off a cliff. His advice? Keep balanced exposure to AI-focused semiconductor names, but be selective.

The GPU Heavyweights

Graphics processing units remain the backbone of the AI infrastructure buildout, and Nvidia Corp. (NVDA) sits firmly at the center of it all.

Pajjuri sees limited threats to Nvidia's full-stack dominance, even as custom Application-Specific Integrated Circuits and rival GPUs make progress. Hyperscaler spending should support Nvidia's order visibility going forward, and the analyst believes the valuation already bakes in some slowdown risk. Translation: there's still upside if things go right.

Advanced Micro Devices, Inc. (AMD) gets credit as a credible secondary GPU supplier, especially after landing a deal with OpenAI. But there's a catch. AMD's valuation already reflects near-term production ramps, which means upside may be limited until later product cycles kick in.

As for Broadcom Inc. (AVGO), Pajjuri acknowledges momentum in custom accelerators but questions whether certain customer programs will last. He also flagged valuation concerns and margin pressures as potential headwinds.

Memory Markets Get a Structural Boost

High-bandwidth memory is turning into a transformative force that could reduce the notorious boom-bust cycles plaguing the memory market. Pajjuri expects HBM to represent roughly a quarter of DRAM industry revenue and grow at over 40% annually through 2028.

Tight supply should keep demand running ahead of supply into 2027, which sets up Micron Technology Inc. (MU) particularly well. The analyst sees persistent DRAM tightness and meaningful earnings upside for Micron as the HBM story plays out.

There's another wrinkle here: GenAI workloads are becoming increasingly memory-intensive, which supports HBM demand. The flip side? Higher memory prices could create headwinds for PCs and smartphones, which are more price-sensitive markets.

Get Applied Materials Alerts

Weekly insights + SMS (optional)

Chip Equipment Makers Stay Busy

Wafer fab equipment spending should maintain an upward trajectory over the next two years, even if China-related spending softens. Pajjuri expects memory-related WFE spending to outpace logic spending, driven by extremely tight supply conditions in memory markets.

The winners here include Lam Research Corp. (LRCX), Applied Materials Inc. (AMAT), and ASML Holding NV (ASML). These names are positioned to benefit as memory manufacturers race to add capacity and meet surging HBM demand.

The Supporting Cast

Beyond the marquee names, Pajjuri highlighted several other beneficiaries of the AI infrastructure buildout.

Marvell Technology Inc. (MRVL) stands out among networking and connectivity plays, while Astera Labs benefits from AI system interconnect trends as data centers get more complex.

Arm Holdings Plc (ARM) gets a nod for its expanding data center footprint and pricing power, which should support long-term growth even as smartphone demand faces memory-related pressures.

The Bottom Line

Pajjuri's overarching thesis is that AI has split the semiconductor sector into clear winners and losers. Selectivity matters now more than ever, especially with valuations sitting above historical averages. Any meaningful slowdown in hyperscale spending could hit stock prices hard, so understanding which companies have durable competitive positions becomes critical.

The message isn't that everything will work out perfectly. It's that the structural drivers remain intact, the spending environment should stay supportive for at least another year or two, and the right names still offer meaningful upside if you're willing to do the work to separate the haves from the have-nots.

NVDA Price Action: Nvidia shares were up 2.96% at $188.57 at the time of publication on Thursday, according to market data.