TMTPOST -- Financial results released on Wednesday showed Nvidia Corporation’ performance was not as bad as Wall Street feared in the face of the tightening export control on artificial intelligence (AI) chips to China.

Credit：Nvidia

Nvidia posted revenue of $44.1 billion for its first fiscal quarter ended April 27, beating analysts expected $43.29 billion. That represented a year-over-year (YoY) increase of 69% in sales, slowing down with a 78% surge three months earlier. On non-GAAP basis, adjusted earnings per share (EPS) shot up 33% YoY to $0.81 after a 71% YoY increase for the preceding quarter, while EPS excluding charge related to H20 chips and tax impact stood at $0.96, still better than estimated $0.93. Adjusted gross margin including H20 charge dropped to 61%, down from 78.9% a year ago. Without H20 charge, the margin was 71.3%, compared with estimates of 71%.

Nvidia said in a statement it incurred a $4.5 billion charge during the April quarter associated with H20 excess inventory and purchase obligations as the demand for H20 diminished under the new requirements for export license. The AI chip giant on April 9 was informed by the U.S. government that a license is required for exports of its H20 products, which were designed primarily for China, into the market.

Sales of H20 chips were $4.6 billion for the first fiscal quarter prior to the new export restrictions, said Nvidia. The company also reported it was unable to ship an additional $2.5 billion of H20 revenue in the quarter due to the restrictions.

Looking forward the second fiscal quarter, Nvidia expected revenue to be $4.5 billion, plus or minus 2%. The outlook is roughly in line with Wall Street forecast as the mid-point is slightly below the consensus expectation of $45.5 billion. The company said its sales guidance reflected a loss in H20 revenue of around $8 billion. Adjusted gross margin is projected to be 71.5% to 72.5%, in line with estimated 71.7%.

The revenue beat suggested demand for Nvidia’s top AI chips remained strong as it said more than half of its customer base is the major tech hyper-scalers. The outlook showcased Nvidia can still generated robust growth despite barriers to revenue potential in China. The Santa Clara, California-based company highlighted some of progress for the April quarter, such as partnership withe OpenAI, G42, Oracle, SoftBank and Cisco in Stargate UAE, a next-generation AI infrastructure cluster in Abu Dhabi, United Arab Emirates.

Nvidia CEO Jensen Huang told analysts on an earnings call the company has some limited options that it is exploring so it can continue to serve the Chinese market, and for now, it has to take a write-off for H20 chips.

“China is one of the world’s largest AI markets and a springboard to global success with half of the world’s AI researchers based there; the platform that wins China is positioned to lead globally today,” Huang said. “However, the $50 billion China market is effectively closed to us. The H20 export ban ended our Hopper data center business in China. We cannot reduce Hopper further to comply.”

Huang cautioned that “shielding Chinese chipmakers from U.S. competition only strengthens them abroad, and weakens American’s position. Export restrictions have spurred China’s innovation and scale.” He continued: “the AI race is not just about chips. It’s about which stack the world runs on.”

For Huang, Washington’s policy has made on on the assumption that China cannot make AI chips, while such assumption “was always questionable, and now its clearly wrong.” “China has enormous manufacturing capability. In the end the platform that wins the AI developers wins AI. Export controls should strengthen U.S. platforms, not drive half of the world’s AI talent to rivals,” said Huang.