By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panel discussed NVDA's potential to reach a $1 trillion run rate by 2027, with mixed views on the likelihood and implications of this achievement. Key concerns include supply constraints, competition, and the risk of slower conversion of orders into revenue. Some panelists are bullish, citing strong demand signals and NVDA's GPU moat, while others are neutral or bearish due to execution risks and potential competition from AMD, Intel, and custom accelerators.
Risk: Supply constraints, particularly TSMC's 3nm wafer capacity and HBM3e supply, as well as slower conversion of orders into revenue.
Opportunity: Reaching a $1 trillion run rate, driven by strong demand signals and NVDA's GPU moat.
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 →
NVIDIA Corporation (NASDAQ:NVDA) is one of the 10 Best AI Stocks to Buy for the Next 10 Years. On March 19, New Street Research added NVIDIA Corporation (NASDAQ:NVDA) to its “best idea list for 2026.” The firm’s analyst, Pierre Ferragu, said that the latest order figures indicate much higher long-term revenue potential than investors are expecting currently.
Ferragu pointed to comments from NVIDIA Corporation’s (NASDAQ:NVDA) CEO, Jensen Huang, at NVIDIA GTC 2026, where Huang said, “I see, through 2027, at least $1 trillion.” New Street Research said that the market’s reaction on this figure being close to expectations was “misplaced.” The research firm said that the company “will likely materially beat 2027 expectations.”
New Street Research noted that Huang had previously stated at GTC Washington in October 2025 that NVIDIA Corporation (NASDAQ:NVDA) had “visibility into $0.5 trillion” of cumulative demand for Blackwell and early Rubin through 2026.
According to Ferragu, the new update shows that NVIDIA Corporation (NASDAQ:NVDA) has “added $500bn of orders since October.” The analyst noted that the company is now on a run rate of more than $1 trillion per year.
NVIDIA Corporation (NASDAQ:NVDA) is an American multinational technology company known for producing graphics processing units (GPUs), AI hardware and software, and high-performance computing (HPC) solutions.
While we acknowledge the potential of NVDA as an investment, we believe certain AI stocks offer greater upside potential and carry less downside risk. If you're looking for an extremely undervalued AI stock that also stands to benefit significantly from Trump-era tariffs and the onshoring trend, see our free report on the best short-term AI stock.
READ NEXT: 11 Best Tech Stocks Under $50 to Buy Now and 10 Best Stocks Under $20 to Buy According to Hedge Funds.
Disclosure: None. Follow Insider Monkey on Google News.
Four leading AI models discuss this article
"The upgrade conflates order visibility with earnings certainty without addressing margin compression, customer concentration, or whether consensus already embedded these figures."
New Street's upgrade hinges on interpreting Huang's '$1 trillion through 2027' as materially above consensus, but the article provides zero evidence of what consensus actually expects. If the Street already priced in $900B+ run rates, this is noise, not signal. The $500B order add since October is real, but orders ≠ revenue (conversion risk, timing risk, customer concentration risk in hyperscalers). Ferragu's claim that NVDA 'will likely materially beat 2027 expectations' is speculative without showing the math. The article also omits gross margin trajectory—if ASPs compress as competition intensifies (AMD, custom silicon), revenue beats don't guarantee earnings beats.
If consensus 2027 revenue is already $950B+, Huang's $1T statement is essentially in-line, and New Street is retrofitting a bullish narrative onto a non-event. Worse: hyperscaler capex cycles are notoriously lumpy and reversible; a single customer pullback could crater orders.
"The market is conflating massive order backlogs with long-term demand sustainability, ignoring the potential for a sharp correction if enterprise AI ROI does not justify the current hardware spend."
New Street’s endorsement of NVDA, predicated on a $1 trillion revenue run rate, assumes an almost frictionless transition from initial AI infrastructure build-out to sustained, high-margin software-driven demand. While the order backlog is undeniably massive, the market is pricing in near-perfect execution. We are moving from a supply-constrained environment to one where the bottleneck shifts to the end-user's ability to monetize these clusters. If enterprise ROI fails to materialize by 2026, NVDA’s valuation—currently trading at a premium reflecting hyper-growth—will face significant multiple compression. The $1 trillion figure is impressive, but it assumes no meaningful competition from custom silicon or a cooling in hyperscaler CapEx.
If the hyperscalers view AI compute as a foundational utility rather than a discretionary expense, NVDA’s revenue could prove to be more resilient and annuity-like than historical semiconductor cycles suggest.
"NVIDIA's headline order run-rate is powerful but only valuable to shareholders if those orders reliably convert to sustained revenue and margins amid capacity, competition, and geopolitical risks."
This is a bullish datapoint but not a free pass. New Street and Jensen Huang highlight an enormous demand signal—claims of a >$1tn run-rate and $500bn added orders since Oct imply material multi-year acceleration for Blackwell/Rubin-driven data center spend. The crucial missing context: is this booked, committed revenue or customer intent? Conversion timing, supply constraints (foundry/GPU capacity, HBM memory), pricing power, and export controls (China) can all materially delay or shrink revenue. Also, much of NVDA’s goodness is already baked into a very rich valuation; the market will punish any sign of slower conversion, margin compression, or incremental competition from AMD/Intel/custom accelerators.
If those order figures are committed and convert on the cadence New Street assumes, NVDA’s earnings and cash flow would justify a meaningful multiple expansion and huge upside—New Street could be understating the upside if demand remains stickier than history suggests.
"NVDA's $500B order surge since October signals revenue potential materially above consensus, driving re-rating if execution holds."
New Street's upgrade highlights NVDA's surging order momentum: $500B added since October 2025, lifting cumulative visibility from $0.5T (Blackwell/Rubin through 2026) to $1T+ through 2027, implying a >$1T annual run rate. Huang's GTC 2026 comments validate hyperscaler capex frenzy, outpacing consensus. With NVDA's GPU moat intact, this supports EPS growth >>20% into 2027, justifying P/E expansion from current ~40x forward levels if ramps execute. Key watch: Rubin production scaling to sustain the trajectory amid power/cooling bottlenecks.
Hyperscalers could slash GPU spend if AI ROI disappoints or custom chips proliferate, turning that $1T visibility into a demand cliff; plus, TSMC concentration exposes NVDA to Taiwan geopolitics.
"TSMC fab capacity, not demand or ROI, is the binding constraint on NVDA's $1T thesis by 2027."
Grok flags TSMC concentration risk—valid. But nobody's quantified the actual constraint. TSMC's 3nm capacity is ~3M wafers/month; NVDA consumes ~15-20% of that already. Rubin scaling to sustain $1T run-rate likely requires 25%+ of TSMC's 3nm output by 2027. That's not a risk—that's a hard ceiling unless TSMC expands dramatically or NVDA diversifies fabs. New Street's upgrade ignores this entirely.
"The $1T run rate is physically constrained by HBM3e memory yields more than by TSMC foundry capacity."
Claude is right about the TSMC ceiling, but both Claude and Grok miss the real bottleneck: HBM3e supply, not just 3nm wafer capacity. Even if TSMC expands, SK Hynix and Micron are struggling to scale high-bandwidth memory yields for Rubin. If NVDA hits a $1T run rate, they need an exponential increase in HBM, which is currently a tighter constraint than foundry logic. This hardware-level dependency makes the $1T target mathematically fragile regardless of hyperscaler demand.
[Unavailable]
"Reported foundry/memory expansions make supply constraints dynamic and NVDA-favorable, not hard ceilings."
Gemini and Claude fixate on static supply ceilings, but TSMC's 3nm/CoWoS expansions (targeting 20%+ YoY growth through 2026, per Q4 earnings) and SK Hynix's HBM4 pilots are explicitly NVDA-aligned. Demand has historically forced rapid scaling—$1T run-rate incentivizes it further, bolstering NVDA pricing power amid constraints. The fragility is overstated; execution risk pales vs. AI ROI uncertainty.
The panel discussed NVDA's potential to reach a $1 trillion run rate by 2027, with mixed views on the likelihood and implications of this achievement. Key concerns include supply constraints, competition, and the risk of slower conversion of orders into revenue. Some panelists are bullish, citing strong demand signals and NVDA's GPU moat, while others are neutral or bearish due to execution risks and potential competition from AMD, Intel, and custom accelerators.
Reaching a $1 trillion run rate, driven by strong demand signals and NVDA's GPU moat.
Supply constraints, particularly TSMC's 3nm wafer capacity and HBM3e supply, as well as slower conversion of orders into revenue.