A flood of new gaming cards for the holidays saturated an already crowded market, but one analyst believes that market is now “on the mend,” with Nvidia Corp. growing more dominant as the rise of artificial intelligence hastens the chip maker’s recovery.
Using survey data from online gaming network Steam’s more than 25 million active users, B of A Securities analyst Vivek Arya noted that 82.6% of users were running Nvidia
NVDA,
cards, a “surprising” jump of about 700 basis points month over month. Those cards range from Nvidia’s GT 10-series running on “Pascal” architecture that were released six years ago to the RTX 40-series “Lovelace” cards that were released in late 2022.
Arya, however, looks upon that data “with caution” given the recent reopening of China’s internet cafes. The analyst noted the survey data showing listed 51.7% of participants using “Simplified Chinese” operating system, versus 26%-27% last month. A similar occurrence of “overcounting” happened in the 2018-2019 cycle, when sales dropped ahead of clearing out of inventory ahead of new launches, he said.
Given that, Arya said he believes the gaming market is “on the mend” following a 50% drop in sales in the fourth quarter, and an expected similar drop in the first quarter as well, which would mark four consecutive quarters of year-over-year declines.
Unlike the 2021’s holiday season, gaming cards from Nvidia, Advanced Micro Devices Inc.
AMD,
and Intel Corp.
INTC,
were readily available and relatively close to their suggested retail price at the end of 2022. Arya has buy ratings on both Nvidia and AMD.
Arya said one aspect of Nvidia’s videogame technology that’s “potentially underappreciated,” is the growing use of AI, which most games already use on a basic level.
Read: Nvidia CEO expects AI revenue to grow from ‘tiny, tiny, tiny’ to ‘quite large’ in the next 12 months
“Existing games today use basics of AI to support computer-controlled responses given certain inputs (ex: computer-controlled opponent), though actions are preprogrammed,” Arya said. “Thus, games/computer-controlled opponents are unable constantly ‘learn’ and adapt from user inputs/prompts.”
“With generative AI, [large language models] can be developed and used to enhance gameplay, enabling anything from creation of new virtual worlds, reducing game/CPU-controlled opponent predictability, and much more,” the B of A analyst noted. “Accelerating various tasks can also possibly speed up game development and reduce costs.”
This post was originally published on Market Watch
