Nvidia forecast quarterly revenue above estimates on Wednesday and announced a stock split, lifting its shares to record-high territory and impressing investors who have tripled the chip maker’s market value in the past year on AI optimism.

Nvidia shares jumped 5.9% to US$1 005 in extended trade, peaking above the psychologically important $1 000 mark and adding about $140-billion in stock market value.

The AI poster child’s stock has surged 90% so far this year, and a close at Wednesday’s after-hours price in the next day’s Wall Street trading session would be a new record high.

The Santa Clara, California-based company said it would split its stock 10-for-one, effective on 7 June. It also said it was raising its quarterly dividend by 150% to 1c/share, on a post-split basis.

“Death, taxes and NVDA beats on earnings. Even in the face of huge expectations, the company once again stepped up and delivered,” said Ryan Detrick, chief market strategist at Carson Group. “The always important data centre revenue was strong, while future revenue was also impressive.”

Wall Street’s main event so far this week, Nvidia’s earnings report could add fresh fuel to a stock market rally that has lifted indexes to record highs this year. Following Nvidia’s results, shares of rival AI-related chip makers AMD and Broadcom each rose about 2%.

Google, Microsoft, Amazon and other technology companies have been competing for a limited supply of Nvidia’s high-end chips as they race to dominate AI computing.

Blackwell

During a conference call with analysts, CEO Jensen Huang said Nvidia’s upcoming Blackwell AI chips will ship in the current fiscal quarter, with production increasing in the following quarter.

Chief financial officer Colette Kress said demand for Blackwell chips could exceed supply “well into next year”.

Nvidia’s contract chip maker, TSMC, has also been working to increase its advanced packaging capacity, a key supply-chain constraint for the processors. The Taiwanese company said in April it expects to more than double its advanced packaging capacity this year.

Nvidia forecast fiscal second-quarter revenue of $28-billion, plus or minus 2%. Analysts on average were expecting revenue of $26.7-billion, according to LSEG data.

First-quarter revenue surged 262% year over year to $26-billion, beating estimates of $24.7-billion. Net income soared 628% to $14.9-billion.

“Demand for Nvidia’s GPU chips remains white-hot,” said Logan Purk, an analyst at Edward Jones. “These results are likely enough to satiate investors’ appetites and reassure the market that AI investment has not seen a slowdown yet.”

Dominating more than 80% of the market for AI chips, Nvidia stands in a unique position as both the largest enabler as well as beneficiary of surging AI development.

Sales at the data centre segment, its largest by revenue, grew 427% to $22.6-billion in the first quarter ended 28 April, coming in above estimates of $21.3-billion, according to data from FactSet.

Among Nvidia’s customers is Meta Platforms, which last month increased the midpoint of its 2024 capital expenditure forecast by about $4-billion.

The high performance of Nvidia’s chips makes them difficult to replace in present AI data centres. Adding to this lead is its proprietary Cuda software framework that developers use to program the AI processors.

While most so-called hyperscalers are also developing their own custom AI chips, analysts do not expect these to eat away at Nvidia’s market share.

Nvidia expects second-quarter adjusted gross margin to be 75.5%, plus or minus 50 basis points. Analysts on average forecast gross margin to be 75.8%.

Nvidia reported first-quarter adjusted gross margin of 78.9% compared with estimates of 77%. Aspiring competitor AMD had recorded an adjusted margin of 52% in its fiscal first quarter. — Arsheeya Bajwa and Stephen Nellis, (c) 2024 Reuters