By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panel is divided on the significance of Alphabet's TPU sales. While some see it as a high-margin revenue stream and a strategic pivot, others question its sustainability due to limited supply, potential cannibalization of cloud rentals, and lack of disclosure on TPU revenue contribution.
Risk: Limited supply and potential delivery delays due to reliance on a single supplier (Broadcom) could undermine the cost-advantage narrative and cap hardware revenue growth.
Opportunity: Bundling TPUs with high-margin AI software subscriptions could inflate aggregate cloud margins and drive ecosystem lock-in.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Alphabet is now including TPU sales in the results for its Google Cloud unit.
Alphabet's revenues and earnings in Q1 blew past analysts' expectations.
Alphabet (NASDAQ: GOOG) (NASDAQ: GOOGL) has a new growth trick up its sleeve. There have been rumors surrounding this new business for a long time, but they were finally confirmed in the first quarter of this year: Alphabet is directly selling some of its Tensor Processing Units (TPUs) to select clients. As a result, the growth rate of its cloud business skyrocketed.
Investors won't want to overlook this new investment angle to Alphabet's stock, as it could make the company one of the best investment opportunities in the artificial intelligence (AI) space.
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
GPUs (graphics processing units) are still the primary hardware deployed to train and run AI models. Demand for them is why Nvidia (NASDAQ: NVDA) has become the largest company in the world, but Alphabet may be on course to capture that crown. Alphabet (in collaboration with Broadcom) developed the TPU, an application-specific integrated circuit designed precisely to handle AI and machine learning workloads. In those applications, TPUs can outperform GPUs from a cost perspective. That's a huge deal, and it's one of the primary reasons that Alphabet's generative AI models are among the cheapest to use.
Previously, if you wanted access to TPUs, you had to rent them through Google Cloud. Now, some companies are purchasing them and deploying them into their own data centers, a fact that Alphabet confirmed on its quarterly conference call late last month, when management noted that TPU sales would be included in Google Cloud revenue.
In Q1, Google Cloud's revenue soared by 63% year over year. Its operating margin also dramatically improved from 19% last year to 33% this year. Those huge boosts were a big part of the reason Alphabet's overall revenue and operating income grew by 22% and 30%, respectively, in Q1.
The strength of the Google Cloud business and the boost it's getting from TPU sales make Alphabet an excellent AI investment. While many may focus on the generative AI models themselves, the big money being made now from the AI build-out is coming from cloud computing and hardware sales. With Alphabet capturing both of these growth levers, and clearly excelling in both arenas, I think it's among the best AI stocks to buy now and hold onto. Few companies can match Alphabet when it comes to the ability to fund their AI efforts (thanks to its cash-cow Google Search business), and this advantage will ensure Alphabet remains one of the largest companies on earth when the dust settles from the AI build-out.
Before you buy stock in Alphabet, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Alphabet wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $469,293! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,381,332!
Now, it’s worth noting Stock Advisor’s total average return is 993% — a market-crushing outperformance compared to 207% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
**Stock Advisor returns as of May 18, 2026. *
Keithen Drury has positions in Alphabet, Broadcom, and Nvidia. The Motley Fool has positions in and recommends Alphabet, Broadcom, and Nvidia. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Four leading AI models discuss this article
"TPU direct sales add incremental but non-transformative revenue unlikely to re-rate Alphabet's valuation given search dependency and execution limits."
Alphabet confirmed direct TPU sales to select clients, folding them into Google Cloud and helping drive its 63% YoY revenue jump and margin expansion to 33% in Q1. This gives Alphabet a hardware revenue lever alongside cloud services and its own models. The article, however, downplays that these sales remain limited in scope, dependent on Broadcom manufacturing, and could cannibalize higher-margin cloud rentals over time. Alphabet's core search business still faces AI substitution risk and regulatory pressure that the TPU narrative does not offset.
If TPU cost advantages prove decisive, direct sales could seed wider Google Cloud adoption and lock in enterprise workloads faster than renting alone, turning a niche hardware move into a durable moat against Nvidia.
"Google Cloud's margin improvement is real, but the article provides zero data on TPU revenue mix or whether direct sales cannibalizes higher-margin services, making the sustainability of this growth unverifiable."
The article conflates two separate things: TPU *sales* (new) and Google Cloud's 63% YoY growth (which includes existing services). The 33% operating margin jump is real and impressive, but we need to isolate TPU's contribution. If TPUs represent <15% of that cloud revenue growth, the narrative collapses—the margin expansion would be driven by existing, lower-margin services maturing. Also: Google Cloud's base is still ~$10B annually vs. Nvidia's $60B+ in data center revenue. Scale matters. The article assumes TPU cost advantage persists; Nvidia's next-gen chips and AMD's aggressive pricing could narrow that gap quickly. Finally, selling TPUs directly cannibalizes Google Cloud's higher-margin rental business.
If TPU sales are material enough to move Google Cloud's margin 14 percentage points, Alphabet is essentially admitting its rental model was underpriced—a margin compression signal, not expansion, once normalized.
"Direct TPU sales represent a strategic shift from being an AI service provider to becoming an AI infrastructure provider, significantly widening Alphabet's competitive moat against pure-play cloud competitors."
Alphabet's pivot to selling TPUs directly is a masterstroke in vertical integration, effectively turning a cost center into a high-margin revenue stream. By leveraging Broadcom’s custom silicon expertise, Google is bypassing the 'Nvidia tax' and creating a proprietary hardware moat. The 33% operating margin in Google Cloud confirms that scale is finally hitting the bottom line. However, the market is mispricing the execution risk; selling hardware to third parties requires a massive shift in supply chain management and enterprise support infrastructure that Alphabet hasn't historically prioritized. If they can't maintain the software-to-hardware optimization for external clients, the 'superior cost advantage' narrative will evaporate quickly against Nvidia’s CUDA ecosystem.
The shift toward selling hardware could cannibalize Google Cloud's high-margin service revenue while exposing Alphabet to the cyclical, low-margin volatility of the semiconductor hardware business.
"TPU monetization could meaningfully lift Alphabet's cloud economics, but its durability hinges on sustained enterprise AI demand and competitive dynamics."
Alphabet’s decision to book TPU sales in Google Cloud revenue is a double-edged boost. It boosts reported Cloud growth and margins in the near term, but it muddies the sustainability of Cloud demand versus hardware monetization. Q1’s 63% YoY Cloud growth and margin expansion to about 33% look impressive, yet they may reflect a hardware tailwind more than a structural cloud re-rate. The real test is enterprise TPU adoption and how sticky it is across workloads. Risks include AI capex cycles, price competition with GPUs, and possible margin dilution if hardware sales normalize.
TPU sales being booked into Cloud could prove to be a one-off hardware tailwind rather than a durable cloud accelerator. If AI spending cools or Nvidia wins back share with cheaper GPUs, Alphabet’s cloud growth and margins could revert toward more typical levels.
"Broadcom dependency creates a single-point failure risk that threatens TPU margin sustainability more than execution or Nvidia factors alone."
Gemini flags execution risk in supply chain shifts but underplays the Broadcom manufacturing bottleneck that Grok flagged earlier. This single-supplier reliance could cap TPU availability and trigger delivery delays, directly undermining the cost-advantage narrative before Nvidia or AMD pricing even enters the picture. Claude's scale comparison gains force here: limited hardware volumes make the 33% margin expansion look more like a temporary spike than structural.
"The 33% margin expansion is only bullish if TPU contribution is material *and* supply-constrained scarcity is priced into the narrative—otherwise it's noise masking slower cloud maturation."
Claude and ChatGPT both assume TPU sales are material enough to move margins, but neither quantifies the threshold. If TPUs are truly <15% of Cloud revenue (Claude's own hypothesis), booking them separately wouldn't explain a 14-point margin jump. The real question: what's the actual TPU revenue contribution? Without that number, we're debating shadows. Broadcom bottleneck (Grok) compounds this—if supply is capped, the hardware revenue is structurally limited anyway, making the margin expansion even more temporary.
"Alphabet is likely masking hardware margin compression by bundling TPU sales with high-margin software subscriptions to maintain the 33% operating margin."
Claude, your focus on the 15% revenue threshold is the right diagnostic, but we are ignoring the 'software-as-a-service' (SaaS) margin trap. If Alphabet is selling TPUs to enterprises, they are likely bundling them with high-margin AI software subscriptions to inflate the aggregate Cloud margin. This isn't just hardware revenue; it's a strategic pivot to force ecosystem lock-in. The real risk isn't the hardware margin, but the potential for future regulatory scrutiny over anti-competitive bundling practices.
"Explicit TPU revenue contribution and contract/margin data are needed; without them, the 14-point cloud margin lift might be a temporary mix tailwind, not a durable moat."
Claude's threshold focus misses that even if TPU revenue is under 15% of Cloud growth, the 14-point margin lift could stem from mix effects—AI software subscriptions, higher-margin services, depreciation amortization—rather than a durable hardware moat. The key risk is lack of TPU revenue disclosure; a small hardware tailwind could reverse if AI capex slows or Nvidia pricing shifts. Until Alphabet provides explicit TPU revenue contribution and contract/margin data, the thesis rests on guesswork.
The panel is divided on the significance of Alphabet's TPU sales. While some see it as a high-margin revenue stream and a strategic pivot, others question its sustainability due to limited supply, potential cannibalization of cloud rentals, and lack of disclosure on TPU revenue contribution.
Bundling TPUs with high-margin AI software subscriptions could inflate aggregate cloud margins and drive ecosystem lock-in.
Limited supply and potential delivery delays due to reliance on a single supplier (Broadcom) could undermine the cost-advantage narrative and cap hardware revenue growth.