By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panel is divided on Cerebras' threat to Nvidia's data-center dominance. While some see potential in Cerebras' 15x inference speedup and OpenAI deal, others caution about execution risks, customer concentration, and wafer-scale manufacturing challenges. Nvidia's entrenched moat, including its CUDA software ecosystem, remains a significant barrier for Cerebras.
Risk: Wafer-scale manufacturing yield and the risk of missing performance SLAs in the OpenAI deal are the most frequently cited concerns.
Opportunity: Cerebras' potential 15x inference speedup and the OpenAI compute deal are seen as significant opportunities.
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 →
Cerebras stock just soared in the double-digits after the biggest IPO of the year.
Nvidia stock has been a surefire winner for investors, surging more than 1,500% in five years.
Nvidia (NASDAQ: NVDA) has been the "go-to" artificial intelligence (AI) stock since the start of this technology revolution. Why? Because the company offers the most powerful chips that fuel the most essential of AI operations, such as training models and guiding them through their work. Other chip designers -- market giants such as Advanced Micro Devices and Intel -- also are benefiting from the AI boom, but Nvidia remains the leader.
Now, however, a young player is set to challenge this market behemoth, and investors are clearly intrigued. Cerebras Systems (NASDAQ: CBRS) began trading on May 14, and the stock soared 68% in the biggest initial public offering of the year. The company raised more than $5.5 billion and finished the day with a market capitalization of nearly $67 billion.
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 »
So, which is the better AI stock to buy after this exciting market debut? Should you go with market giant Nvidia or promising challenger Cerebras? Let's find out.
Nvidia has been a surefire winner for investors over the past several years -- for example, over five years, it's soared 1,500%. This is due to its dominance in the AI chip market and the fact that this has translated into explosive earnings growth. Nvidia makes the graphics processing units (GPUs) that, year after year, have proven themselves to be the most powerful. So, tech giants eager to offer themselves the best AI tools have piled into Nvidia's systems.
As a result, Nvidia's revenue climbed to $215 billion in the recent full year -- up from $27 billion just three years ago. And this is at a high level of profitability, with gross margin exceeding 70%.
The big question investors have posed is whether chip rivals will eventually chip away (excuse the pun) at Nvidia's market share. So far, Nvidia's line of defense has been its commitment to innovation as well as its deep presence across industries. The company has updated its chips and systems on an annual basis, making it difficult for rivals to surpass its technological strengths. Nvidia also has forged partnerships with a broad range of companies to establish itself as a key player in AI development across industries -- for example, in a partnership with Nokia, it's bringing AI to the launch of 6G and future telecom systems.
All of this may help Nvidia maintain its market dominance and continue to deliver enormous growth as the AI boom marches on.
Cerebras is the exciting young player that already has scored an AI win for some investors -- those who participated in the IPO. Now, those early investors and others who buy the stock in the coming days hope this is just the beginning of a new AI success story.
The company could have what it takes to make that happen. Cerebras offers AI chips that are much bigger than those of Nvidia -- and the company says this results in impressive speed. In inference, or the thinking process models go through, Cerebras chips have generated answers as much as 15 times faster than "leading GPU-based solutions."
"We believe that our high-speed AI solutions give us a meaningful competitive advantage in this market," the company wrote in its prospectus.
This has translated into revenue growth, with revenue of $24 million in 2022 increasing to more than $510 million last year. And this momentum may continue as customers turn to this newish player. Cerebras in January announced a $20 billion deal, spanning several years, to offer compute to AI lab OpenAI.
Cerebras offers its customers hardware on site, or they may access compute via the Cerebras Cloud or through other cloud service providers, such as Amazon's Amazon Web Services (AWS).
So growth may just be getting started for Cerebras -- and its investors.
Both of these companies may benefit from the AI story as it advances. Demand remains high for compute, and the AI market is set to reach into the trillions of dollars in the coming years. It's also important to note that the more AI is applied to real-world problems, the more customers will need computing power.
Which stock is the better buy today? Though Cerebras might have a very bright future, after its massive initial gain, I would wait for a dip to get in on the stock. It's important to remember that Cerebras is in the early days of its AI story and doesn't offer the earnings strength or market leadership of giant Nvidia -- this younger player still carries more risk as it works toward generating consistent operating profit and grows its AI presence.
Of course, Nvidia stock has climbed significantly over time. But its long-term prospects and valuation today suggest it has plenty of room to run. Right now, Nvidia stock looks reasonably priced at 28x forward earnings estimates, down from more than 40x just a few months ago, offering growth investors a buying opportunity they may not want to miss.
Before you buy stock in Nvidia, 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 Nvidia 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 17, 2026. *
Adria Cimino has positions in Amazon. The Motley Fool has positions in and recommends Advanced Micro Devices, Amazon, Intel, 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
"Cerebras' OpenAI contract and inference speed edge create a credible path to share gains that Nvidia's valuation does not yet price in."
Cerebras' $20 billion multi-year OpenAI contract and 15x inference speedup claims represent a direct threat to Nvidia's data-center moat that the article underplays. While NVDA's $215 billion revenue and 70%+ gross margins reflect current dominance, Cerebras' wafer-scale chips target exactly the inference workloads now driving incremental spend. At a post-IPO $67 billion valuation after a 68% pop, CBRS still trades at a fraction of Nvidia's scale yet with faster early revenue ramp from $24 million to $510 million. Investors should watch whether OpenAI migrates production workloads before assuming Nvidia's 28x forward multiple is safe.
Cerebras has never posted consistent operating profit and its $20 billion OpenAI deal may prove back-end loaded or contingent, leaving it vulnerable to Nvidia's annual chip cadence and CUDA ecosystem lock-in.
"Cerebras' IPO valuation reflects speculative bubble pricing on unproven technology and unmonetized claims, not comparable to Nvidia's cash-generative dominance."
The article frames this as Nvidia vs. Cerebras, but that's a false binary. Cerebras raised $5.5B at a $67B valuation on $510M revenue — that's 131x sales, versus Nvidia at roughly 12x. The IPO pop (68%) is classic euphoria pricing, not fundamental validation. Cerebras claims 15x inference speedup, but inference is lower-margin than training, and the claim lacks independent verification. More critically: Cerebras has zero operating profit, OpenAI's $20B deal is multi-year (not annual), and competing against Nvidia's ecosystem lock-in is brutally hard. The article's Nvidia valuation (28x forward P/E) is presented as 'reasonably priced,' but omits that Nvidia faces margin compression risk as competition intensifies and GPU oversupply risks emerge in 2025-26.
Cerebras could genuinely own inference at scale if its architecture proves superior in production, and a $67B valuation might be cheap if it captures even 5-10% of a $500B+ inference market by 2027.
"Nvidia’s true competitive advantage is the CUDA software ecosystem, not just raw chip performance, making it significantly harder for Cerebras to capture market share than the article implies."
The article presents a classic 'incumbent vs. disruptor' narrative, but it ignores the fundamental architectural divergence between Nvidia’s general-purpose GPU clusters and Cerebras’s Wafer-Scale Engine (WSE). Nvidia’s moat isn't just hardware; it’s the CUDA software ecosystem, which creates massive switching costs for developers. Cerebras is betting on specialized performance for specific workloads, which is high-risk, high-reward. While Nvidia’s forward P/E of 28x looks attractive, it assumes continued 20%+ EPS growth. If hyperscalers like Microsoft or Google succeed in developing their own custom silicon (ASICs), Nvidia’s margins will face structural compression that the article’s 'buy' thesis completely glosses over.
Nvidia’s reliance on TSMC for manufacturing creates a single point of failure, whereas Cerebras’s unique wafer-scale approach could theoretically decouple from standard GPU supply chain bottlenecks if they achieve better yields.
"Nvidia's moat from software, ecosystem, and scale justifies staying long NVDA even as Cerebras remains a speculative wildcard."
The article treats Cerebras as a mere challenger while Nvidia's dominance seems secure. The strongest counterpoint is that the AI hardware cycle could endure longer for Nvidia thanks to CUDA software moat, broad data-center exposure, and regular chip upgrades. Cerebras' claim of 15x faster inference and the reported $20B OpenAI compute deal are meaningful, but come with execution risk, customer concentration, and unclear profitability path. If AI capex stays robust, Nvidia can re-rate on growth visibility; but a macro slowdown or AI cooldown could compress its multiple. Cerebras is a high-risk, high-reward wildcard—not a substitute for Nvidia's entrenched moat.
Cerebras' performance edge could surprise if adoption scales beyond a marquee deal, potentially re-rating CBRS and challenging Nvidia's moat more quickly than expected.
"Cerebras adoption faces a multi-year lag from ecosystem barriers, giving Nvidia time to protect margins via chip cadence."
Claude's 131x sales multiple critique underplays the phased rollout risk Gemini raised around CUDA switching costs. Even a verified 15x inference edge from Cerebras would require OpenAI to rewrite kernels and validate at scale, pushing meaningful revenue share beyond 2026. Nvidia can defend its 28x forward P/E by accelerating Blackwell releases during that window, while Cerebras margins may compress if it discounts to win non-OpenAI workloads.
"Cerebras' wafer-scale architecture creates a single-point-of-failure yield risk that Nvidia's multi-node strategy structurally avoids."
Grok's 2026 timeline assumes Cerebras executes flawlessly on kernel rewrites and validation—a massive assumption. But he's underselling the real risk: OpenAI's $20B deal likely includes performance SLAs. If Cerebras misses those, OpenAI walks back to Nvidia regardless of theoretical speedup. Meanwhile, nobody's addressed yield. Wafer-scale manufacturing is notoriously fragile. One bad fab quarter tanks Cerebras's entire supply chain, while Nvidia spreads risk across multiple TSMC nodes. That's the execution moat Claude and Gemini both missed.
"Cerebras faces a fundamental manufacturing yield risk that makes their current valuation unsustainable relative to Nvidia's proven economies of scale."
Claude, you’re hitting on the critical yield risk, but you’re overlooking the capital intensity. Wafer-scale isn't just fragile; it’s an absolute cash incinerator. Cerebras’s $510M revenue against a $67B valuation assumes they can scale manufacturing without the massive economies of scale Nvidia enjoys. If they hit a yield wall, their unit economics collapse, and the $20B OpenAI deal becomes a liability, not an asset. They aren't just fighting Nvidia; they are fighting the laws of silicon manufacturing.
"Yield concerns don't automatically imply failure; diversification and proof of reliability could unlock value, keeping Cerebras investable even if OpenAI remains a focal point."
Gemini overstates wafer-scale fragility by focusing on yield as doom. Even with high capex, Cerebras could de-risk via multiple revenue streams (licensing IP, managed services) beyond a single OpenAI deal, reducing unit economics pressure if adoption scales. The bigger risk remains OpenAI contract concentration and execution risk; if Cerebras diversifies wins and proves reliability, the 67B valuation could be justified despite yield concerns.
The panel is divided on Cerebras' threat to Nvidia's data-center dominance. While some see potential in Cerebras' 15x inference speedup and OpenAI deal, others caution about execution risks, customer concentration, and wafer-scale manufacturing challenges. Nvidia's entrenched moat, including its CUDA software ecosystem, remains a significant barrier for Cerebras.
Cerebras' potential 15x inference speedup and the OpenAI compute deal are seen as significant opportunities.
Wafer-scale manufacturing yield and the risk of missing performance SLAs in the OpenAI deal are the most frequently cited concerns.