LIVE: Nvidia beats Q1 2027 earnings as revenue surges 85% to $81.6 billion

Chinese companies including Alibaba (BABA), Tencent (TCEHY), ByteDance, and JD.com (JD) received approval from the U.S. Commerce Department last week to buy Nvidia’s H200 chips, as Cryptopolitan previously reported.

But a U.S. trade representative said chip export controls were not discussed during last week’s China talks. That means any broader loosening of restrictions around H200 sales may still be far away.

Nvidia is also moving fast to build out its supply chain as it prepares for a much larger AI market. Jensen Huang, Nvidia’s CEO, said the idea of Nvidia becoming many times bigger is not impossible.

Jensen said the company is investing heavily across what he called the AI industry’s “five-layer cake.” That includes energy, chips, infrastructure, models, and applications.

He said Nvidia’s top use of its growing cash pile is to support suppliers as demand keeps rising. Jensen said the company is growing by hundreds of billions of dollars at a time, so it needs suppliers to have enough capacity to keep up with that growth.

Jensen Huang, Nvidia’s CEO, said the company has now largely given up China’s AI chip market to Huawei as U.S. export rules keep blocking Nvidia from selling its most advanced processors into the country.

Jensen was interviewed live by CNBC right after he got done with the earnings call, and he them that demand in China is still huge, but Nvidia is no longer in a strong position there.

He said Huawei is powerful, had a record year, and is likely heading for another very strong year as local Chinese chip companies benefit from Nvidia’s exit.

Jensen said Nvidia has basically left that market open for Huawei and China’s domestic chip ecosystem.

China used to be a major piece of Nvidia’s Data Center business. The market once made up at least 20% of the company’s Data Center revenue, but Nvidia has been mostly locked out since the Trump administration told the company in April that it would need a license to ship chips to China and a few other countries.

Jensen also sounded careful about any quick return to China. He said Nvidia has told analysts and investors to build their forecasts around zero China approvals for advanced chip sales.

He said that is why Nvidia’s guidance and expectations do not count on China reopening soon.

Still, Jensen made clear that Nvidia wants to go back if the rules change. He said the company would be glad to serve Chinese customers again, adding that Nvidia has many customers and partners there and has operated in the country for 30 years.

Jensen Huang is still talking in the earnings call even, he just called the first quarter “extraordinary,” pointing to surging demand for AI computing.

Jensen said demand has moved sharply higher because agentic AI is now here. He said AI is no longer just producing demos or simple outputs, but is now doing useful work that creates value for companies.

He said Nvidia is positioned at the center of that shift because the company spent more than 30 years building its compute platform before this cycle arrived. Jensen said Nvidia had built for this moment before agentic AI became a major market driver.

Jensen also explained why Nvidia is changing how it reports its business. He said the new structure is meant to help investors understand the company more clearly.

He said AI now sits at the center of Nvidia’s business, but the customers using Nvidia’s technology are not all the same. Large cloud companies such as Meta (META) and Alphabet (GOOGL) are buying huge amounts of computing equipment and GPUs for major data center projects.

Governments, smaller companies, and businesses working on advanced robotics have different needs. Jensen said those groups use different technology stacks, run on different operating systems, and buy Nvidia’s products in different ways.

He said the new reporting setup is the simplest way to show how the business works across those separate markets.

Nvidia’s earnings call is no over.

Jensen Huang, Nvidia’s CEO, said during the Q&A that the company is winning more business in inference computing, which is when an AI model is used in real time after it has already been trained.

Jensen said Nvidia’s position is also growing because of physical AI, a field the company has worked on for years.

Physical AI covers systems that deal with the real world, including self-driving cars, factory robots, and other machines that need AI to make decisions outside normal software.

He said Nvidia is nearly alone in serving that market today, and that its share of inference is rising quickly as demand expands.

Jensen was also asked about Nvidia’s $1 trillion chip revenue forecast. He said that number does not include standalone sales from Vera CPUs.

He added that Vera is expected to be Nvidia’s biggest revenue source outside that $1 trillion forecast.

Colette Kress, Nvidia’s chief financial officer, said on the earnings call that the first quarter was “exceptional” after the company set new highs for revenue, operating income, and free cash flow.

Colette said spending on AI infrastructure is still rising fast and could reach $3 trillion to $4 trillion a year by the end of the decade.

She said AI has moved from being optional to being needed across industries, as companies use it to improve productivity in different jobs and business areas.

Colette also said Nvidia’s Blackwell architecture is now widely used across the market. She said it has been adopted and deployed by every major hyperscaler, every cloud provider, and every major model maker.

She also pointed to Nvidia’s Vera CPU as a new growth driver. Colette said Vera gives Nvidia access to a $200 billion CPU market, with the company already seeing a path to almost $20 billion in CPU revenue this year.

Colette said Nvidia still has full confidence in its forecast for $1 trillion in revenue tied to the company’s Blackwell and Rubin architectures.

Nvidia (NVDA) said Wednesday in a regulatory filing that it has not booked any revenue from the U.S. government’s licensing program for its H200 AI chips.

The company also said it still does not know whether any shipments of those processors will be cleared for China.

The company has not received approval from Chinese officials. Beijing has been pushing harder to support its own chip suppliers, and Nvidia said it was basically blocked from competing in China’s data center compute market during the first quarter.

Jensen Huang, Nvidia’s CEO, spoke Monday after joining Trump on a trip to China last week. Jensen said he believed China’s market would eventually open again to U.S. chipmakers.

That has not happened yet. Trump’s talks with Chinese President Xi Jinping did not produce an immediate path for Nvidia to sell H200 processors into China.

The issue remains politically sensitive in Washington. Some China hawks worry that Beijing could use advanced U.S. AI chips to strengthen its military.

The Trump administration has argued the opposite case, saying controlled sales could slow down Chinese rivals, including heavily sanctioned Huawei, by making it harder for them to gain ground against U.S. AI chip designers.

Nvidia’s earnings call has now started. Cryptopolitan is in the room, and we will bring updates from Nvidia executives as they speak.

Nvidia (NVDA) gave investors a stronger second-quarter forecast after reporting another record quarter.

For Q2 fiscal 2027, the company expects revenue of $91.0 billion, plus or minus 2%. Nvidia said that forecast does not include any Data Center compute revenue from China.

The company expects GAAP gross margin of 74.9% and non-GAAP gross margin of 75.0%, plus or minus 50 basis points. It also expects GAAP operating expenses of about $8.5 billion and non-GAAP operating expenses of about $8.3 billion.

For the full 2027 fiscal year, Nvidia expects both GAAP and non-GAAP tax rates to land between 16.0% and 18.0%. That excludes one-time items and any major changes to the company’s tax environment.

The first-quarter numbers showed how much bigger the business has become. GAAP revenue came in at $81.615 billion, up 20% from $68.127 billion in Q4 FY26 and up 85% from $44.062 billion in Q1 FY26.

GAAP gross margin was 74.9%, down 0.1 percentage points from the prior quarter but up 14.4 percentage points from last year. GAAP operating expenses rose to $7.621 billion, up 12% quarter over quarter and 52% year over year.

GAAP operating income reached $53.536 billion, up 21% from the previous quarter and 147% from a year ago. GAAP net income climbed to $58.321 billion, rising 36% sequentially and 211% from last year. GAAP diluted earnings per share were $2.39, up 36% from Q4 and 214% from Q1 FY26.

On a non-GAAP basis, revenue was also $81.615 billion. Non-GAAP gross margin was 75.0%, compared with 75.1% in the prior quarter and 60.8% a year earlier.

Non-GAAP operating expenses came in at $7.449 billion, up 12% from Q4 and 49% from last year. Non-GAAP operating income reached $53.783 billion, up 21% quarter over quarter and 147% year over year.

Non-GAAP net income was $45.548 billion, up 17% from $38.969 billion in Q4 and up 139% from $19.094 billion last year. Non-GAAP diluted earnings per share came in at $1.87, up 18% from the previous quarter and 140% from the year-ago period.

Nvidia shares rose 1.3% before the close, but moved 0.6% lower after hours following the report.

Nvidia (NVDA) reported record first-quarter revenue of $81.6 billion for the period ended April 26, 2026, beating the level Wall Street had been watching before the print. Revenue rose 20% from the previous quarter and jumped 85% from the same period last year.

Margins stayed strong. GAAP gross margin came in at 74.9%, while non-GAAP gross margin was 75.0%. GAAP diluted earnings per share were $2.39, and non-GAAP diluted earnings per share came in at $1.87.

Jensen Huang, founder and CEO of Nvidia, said the buildout of AI factories is speeding up fast. Jensen said agentic AI is now being used for real work across companies and industries, and he said Nvidia sits at the center of that shift because its platform runs across clouds, AI models, data centers, and edge systems.

Nvidia also returned a record $20.0 billion to shareholders during the quarter through share buybacks and cash dividends. The company ended the quarter with $38.5 billion left under its repurchase plan.

On May 18, 2026, Nvidia’s board approved another $80.0 billion for share repurchases, with no expiration date. The company also raised its quarterly cash dividend from $0.01 per share to $0.25 per share. The dividend will be paid on June 26, 2026, to shareholders on record as of June 4, 2026.

Nvidia is also changing how it reports its business. The company will now split its market platforms into Data Center and Edge Computing. Inside Data Center, it will report Hyperscale and ACIE, which includes AI Clouds, Industrial, and Enterprise.

Hyperscale will cover revenue from public cloud providers and the largest consumer internet companies. ACIE will track Nvidia’s opportunity in AI-focused data centers and AI factories across industries and countries. Edge Computing will include devices used for agentic AI and physical AI, including PCs, game consoles, workstations, AI-RAN base stations, robotics, and automotive.

Under Nvidia’s old reporting categories, Data Center compute revenue reached a record $60.4 billion, up 77% from a year earlier and 18% from the prior quarter. Data Center networking revenue also hit a record, rising to $14.8 billion, up 199% year over year and 35% sequentially.

Stocks head to close higher on Wednesday as traders reacted to lower oil prices, softer Treasury yields, and new hopes that Trump’s war in Iran could be nearing a resolution.

The Dow Jones Industrial Average climbed 640 points, or 1.3%. The S&P 500 gained 1%, while the Nasdaq Composite rose 1.4%.

Oil prices fell sharply during the session. West Texas Intermediate crude futures dropped 5.66% to settle at $98.26 per barrel, while Brent crude fell 5.63% to close at $105.02 per barrel.

The move came after President Donald Trump told reporters that his administration was in the “final stages” of talks with Iran. That eased some of the pressure that had built across markets from fears of a wider conflict and higher energy costs.

Bond yields also moved lower. The 10-year Treasury yield fell more than 8 basis points on Wednesday, while the 30-year yield dropped 6 basis points. One basis point equals 0.01%.

Rates had shaken investors in recent days. The 30-year Treasury yield reached its highest level since 2007, and the 10-year yield moved close to multi-year highs.

The concern is that higher oil prices could bring inflation back into focus while the Federal Reserve prepares to be led by Kevin Warsh.

Investors are worried that the Fed may already be behind in fighting inflation, and that rising rates could put more pressure on an economy already dealing with expensive energy.

The U.S. dollar has flashed a golden cross, a technical signal that Bank of America Securities says could point to more upside through the summer.

A golden cross happens when a short-term moving average moves above a longer-term one. For the U.S. Dollar Index, the signal appeared on April 7, when the index closed at 99.86. The index tracks the dollar against a basket of six major currencies.

Paul Ciana, technical strategist at Bank of America Securities, said the pattern has usually leaned bullish in past market cycles.

Since 1971, the same signal has appeared 39 times, Paul wrote. In those cases, the Dollar Index was higher 20 to 70 trading days later about 66% to 79% of the time. That points to a possible window between May and July.

Currency traders are also dealing with calmer price action. Volatility in the U.S. dollar has dropped sharply, pushing investors in the $9.5 trillion-a-day foreign exchange market toward carry trades and relative value bets as they search for returns.

The dollar is roughly flat against its major peers this year. The JPMorgan Global FX Volatility Index, a broad gauge of currency swings, is now close to its lowest level since 2024.

Short sellers are not backing away from chip stocks, even after getting crushed for most of the past year by the AI rally.

The setup matters because Nvidia’s results have become a major signal for the entire AI trade. A strong forecast could quickly pull buyers back into semiconductor names, but bears are still holding on.

Qualcomm is one of the clearest examples. Bearish bets against the company recently stood near $11.8 billion in notional value, the highest level in at least 10 years, based on data from S3 Partners.

The last major short build-up was in 2018, when bets against the stock moved close to $10 billion during the period around Qualcomm’s failed attempt to buy NXP Semiconductors (NXPI).

Micron is seeing the same kind of pressure. Short interest in the stock is sitting just below a 52-week high. Bearish positioning in Nvidia and Intel (INTC) is also still higher than it was at the start of the year, even though it has come down from recent highs.

The positioning shows that some investors are questioning how much more semiconductor valuations can stretch after the AI boom lifted the sector for years.

Chip companies have been a major force behind the broader market rally, as investors kept buying firms tied to AI infrastructure spending.

That trade has been dangerous for bears. Many short sellers have already been forced to close positions after AI hardware demand stayed stronger than expected and chipmakers kept delivering solid numbers.

The PHLX Semiconductor Index is still up more than 65% year to date, even after posting its first weekly loss in seven weeks.