Nvidia is spearheading the AI industry’s growth with its latest data centers focused on developing and deploying models for the financial services industry.  

AI factories are among a new product class that can help organizations use AI models at lower costs, higher efficiency rates and faster time to market, Malcolm deMayo, global vice president of financial services at the chip-making giant, told Bank Automation News.

Nvidia’s AI factories are data centers where the entire tech stack, including processing units and software, are built in-house, deMayo said.  

“Yesterday’s data centers or the current data centers are these massive complexes, built primarily to retrieve information that’s been generated through transactions or through communications,” deMayo said. “An AI factory can generate content [from the data collected] based on the relevance and provide you with a contextually relevant answer.” 

Any industry that deals with tremendous amounts of data and requires high computational power will be a great customer for Nvidia’s AI factories, Raj Joshi, senior vice president for tech, media and telco at Moody’s Ratings, told BAN. 

“If you look at the share of tech spending by industry verticals, finance has been always the first to adopt technologies and finance is one of the largest verticals of tech spending,” Joshi said. “They are going to be large users of AI.” 

AI factories can help complex financial institutions like insurance companies and global banks — organizations with many business verticals like wealth management, consumer and business banking — improve operational efficiency and keep sensitive data safe, Joshi said. 

“For smaller financial institutions, they don’t have to create their own infrastructure,” Joshi said. “For larger companies, it might be quick time to market.” 

Inside the data center 

The AI factory’s tech stack, built entirely with Nvidia’s hardware and software, has more computational power and simplified onboarding, deMayo said, adding that users can experience performance and processing gains of up to 10 times. 

The Santa Clara, Calif.-based company provides its AI factory models through public clouds like Amazon Web Services, Google Cloud and Microsoft Azure, or it can be accessed on-premises, deMayo said. 

Businesses can also choose what solutions they want from the AI factory as Nvidia’s offerings are highly customizable and can work with almost any tech architecture, deMayo said. 

The chipmaker expects to continue building offerings for AI-related workloads and has been pouring billions of dollars into research and development, deMayo said.  

“Every year our R&D investment continues to climb and it’s very focused in developing solutions to problems in AI and data processing,” deMayo said. 

AI factories for financial services 

Major financial institutions including BNY Mellon are using Nvidia’s AI factories “to build AI literacy, to run experiments, to build AIs that are going to transform banking,” deMayo said. 

While Nvidia doesn’t break down its revenue by economic sector, financial services is a major contributor and the company expects it to grow, deMayo said. 

“I can tell you that the leading firms in financial services across segments are building AI factories,” deMayo said. “Some banks are doing this in cloud, some banks will do it on-premises and some will do both.” 

Nvidia’s stock jumped 6.76% today and closed the trading session at $127.4, according to Nasdaq. 

[stock_market_widget type=”inline” template=”generic” assets=”NVDA” markup=”{name} ({symbol}) is trading at {price} ({change_pct}) as of {last_update}” realtime=”true” api=”yahoo-finance”].

Early-bird registration is now available for the inaugural Bank Automation Summit Europe 2024 in Frankfurt, Germany, on Oct. 7-8! Discover the latest advancements in AI and automation in banking. Register now.