Investors worried about Micron Technology’s opportunity in high-bandwidth memory were looking for company CFO Mark Murphy to project confidence at a Wolfe Research conference on Wednesday — and he looks to have delivered.

Murphy addressed what he called “inaccurate reporting” about the company’s next-generation high-bandwidth-memory positioning, saying Wednesday morning that the company’s HBM4 offering is in high-volume production. Shipments to customers have “successfully” ramped in the first quarter, he noted, with that development occurinng a year ahead of when the company was initially expecting.

Murphy said that Micron’s HBM yield is on track, referring to the amount of usable chips that come from a single wafer.

“We’re highly confident in our HBM4 product performance and quality and reliability,” he said.

Micron’s stock shot about 10% higher on Wednesday.

Shares fell earlier this week after reports that Micron’s HBM4 rival, Samsung Electronics, will start mass production of its HBM4 later this month for use in Nvidia’s upcoming Vera Rubin graphics processing units. That news coincided with a report from research firm SemiAnalysis that said Micron’s HBM4 is unlikely to be part of the supply for Rubin’s first 12 months, as Micron has been unable to meet Nvidia’s pin-speed requirements.

Looking ahead, Murphy said the company “is in the best competitive position it’s ever been in, at a perfect time.”

Most of the demand is being driven by the artificial-intelligence buildout, the CFO said, and he sees demand getting even stronger now that hyperscalers including Alphabet’s Google and Amazon Web Services just raised capital-spending forecasts in their recent earnings reports. As AI models get larger and more complex, memory is becoming a more crucial component to support longer context windows, or the amount of data a model is processing at one time.

The AI-driven demand has led to intense supply shortages in the memory and storage markets, however — a dynamic that Murphy noted has helped the company’s “positive financial performance.” Micron and other memory and storage companies at the center of the demand boom have been able to raise prices on products due to the supply tightness. At the same time, the companies have been reluctant to add more capacity out of fear of ending up in an oversupply if demand slows down.

“We just simply do not have enough supply to meet demand, and by a substantial margin,” Murphy said. He added that the company is investing in ways to keep up supply, including in greenfield capacity that is expected to come online starting in the middle of next year.

Late last month, Micron announced plans for a new advanced wafer-fabrication facility at its complex in Singapore for manufacturing NAND, a a type of memory that has also seen surging demand from its importance for inference, or the process of running AI models in data-centers.

Micron is “in a favorable price environment” from supply constraints, Murphy said. While the company isn’t quantifiably lifting its financial outlook, he said that Micron has a more upbeat view than it did when it reported earnings back in December.

Morgan Stanley analyst Joseph Moore on Wednesday noted that Micron looks poised to continue benefiting from soaring memory prices. Current market conditions suggest Micron could come in “at or above” the higher end of its revenue guidance for the fiscal second quarter, Moore added. In December, Micron guided for revenue of $18.7 billion, plus or minus $400 million, for its second quarter — far above the FactSet analyst consensus for $14.3 billion.

On that, Moore said Micron’s outlook translates to about 30% sequential growth in average sales prices for dynamic random-access memory and NAND.