2 AI Stocks That Could Go Parabolic

Wall Street may argue all about the artificial intelligence (AI) bubble, but big tech’s spending plans in global AI infrastructure are too large to ignore, right now. Amazon, Alphabet, Microsoft, and Meta Platforms expect to collectively spend roughly $650 billion on AI infrastructure in 2026, up about 60% year over year.

Image source: Getty Images.

With a significant portion of this spending already locked in through data center permits, power contracts, and supply agreements, core AI infrastructure players such as Nvidia (NVDA 1.69%) and Broadcom (AVGO 1.45%) are likely to benefit from the near-term demand.

Here’s why stocks of these two companies can go parabolic in the coming years.

Nvidia

Nvidia’s CEO, Jensen Huang, refers to AI as the “largest infrastructure buildout in human history.” The company estimates global AI infrastructure spending to reach $3 trillion to $4 trillion by the end of 2030. With a full-stack AI platform that includes hardware, software, and enterprise platforms and an annual product cadence (releasing new upgrades and architectures each year), the company is well-positioned to capitalize on this opportunity.

Today’s Change

(-1.69%) $-3.15

Current Price

$183.79

Key Data Points

Market Cap

$4.5T

Day’s Range

$182.18 – $187.55

52wk Range

$86.62 – $212.19

Volume

3.9M

Avg Vol

181M

Gross Margin

70.05%

Dividend Yield

0.02%

In 2026, hyperscalers are increasingly adopting Nvidia’s Blackwell B200 systems, which the company claims deliver up to 30x higher inference (real-time model deployment) performance and better power efficiency than the prior-generation H100 chips. The company is also focusing on selling rack-scale solutions such as GB200 NVL72 “AI factory” systems, which bundle CPUs, GPUs, networking interconnects, and supporting software into complete AI server solutions, instead of stand-alone chips. This transition can eventually boost average selling prices and the company’s margins.

Beyond Blackwell, the company has introduced the Vera Rubin system, a six-chip platform designed to deliver even higher performance and power efficiency than Blackwell systems for complex AI and agentic AI workloads. With Vera Rubin currently in full production, the company expects to launch these systems in the second half of 2026. Management has highlighted cumulative revenue visibility for Nvidia’s Blackwell and Rubin systems of over $500 billion from the start of 2025 through the end of 2026.

Considering these tailwinds, Nvidia’s stock may soar in the coming years.

Broadcom

Broadcom is also a major beneficiary of the increasing enterprise and hyperscaler spending on custom chips and high-speed networking in AI data centers.

According to research firm Counterpoint, hyperscalers are increasingly deploying custom AI server chips for specialized AI training and real-time deployment workloads. Broadcom is estimated to account for almost 60% share of this market by 2027.

Today’s Change

(-1.45%) $-4.79

Current Price

$326.38

Key Data Points

Market Cap

$1.6T

Day’s Range

$325.42 – $334.15

52wk Range

$138.10 – $414.61

Volume

535K

Avg Vol

31M

Gross Margin

64.71%

Dividend Yield

0.73%

That growing adoption is already translating into a multiyear backlog. Broadcom exited fiscal 2025 (ending Nov. 2, 2025) with a consolidated backlog of $162 billion, of which custom AI chips accounted for $73 billion to be delivered over the next 18 months. The company’s networking switches are also seeing increased demand in AI data centers, as evidenced by the over $10 billion AI switch backlog at the end of fiscal 2025.

Broadcom reported a 65% year-over-year jump in AI revenues to $20 billion, driven by demand for custom chips, high-speed networking, and optical interconnects. This revenue growth trajectory can become even steeper if the company’s backlog converts as expected or faster, and it has a few new multiyear client wins.

Hence, Broadcom can see parabolic growth even at current elevated valuation levels.