How NVIDIA Increased Data Center Revenue Ninefold in Three Years

NVIDIA (NVDA) is a well-known name in the gaming and data center markets. Gaming is its biggest business, whereas data center is its fastest-growing business. Its data center business rose almost nine-fold from just $340 million in 2015 to $3 billion in 2018. This growth was driven by an increasing number of applications using its GPUs (graphics processing units). NVIDIA is a leader in deep learning and is slowly gaining traction in inference, where CPUs (central processing units) dominated.

Ready to put your morning scrolling to use? Sign up for Bagels & Stox, our witty take on the top market and investment news straight to your inbox! Whether you’re a serious investor or just want to be informed, Bagels & Stox will be your favorite email.

NVIDIA’s data center business witnessed a temporary pause in revenue growth in the fourth quarter of fiscal 2019 as some large hyperscalers moved to absorb the big purchases they made in early fiscal 2019. However, the growth is expected to resume in the second half of this year.

Intel (INTC), which owns more than 95% of the server CPU market, also acknowledged NVIDIA as its data center competitor. Intel’s $21 billion server CPU business is seven times the size of NVIDIA’s data center business.

At the 2019 Investor Day, NVIDIA’s CEO, Jensen Huang, stated that the company moved from an accelerator company, which accelerates a particular function, to an accelerated computing platform, which accelerates a domain of applications. He compared its platform to Intel’s x86 CPU (central processing). Just like x86 can run all types of software, NVIDIA’s GPU can accelerate all types of applications and domains.

During the 2019 Investor Day, NVIDIA’s executive vice president of Worldwide Field Operations, Jay Puri, talked about the company’s go-to-market strategy and the value proposition, which has been driving its data center sales. He explained that NVIDIA’s GPUs speed up computing by 25x for an HPC (high-performance computing) workload and even 100x for an AI (artificial intelligence) training workload as compared to CPUs. NVIDIA’s GPUs deliver such performance while reducing the server count from 5,000 to just 200 and in some cases less than 50.

A lower server count reduces the space and power requirement to keep servers, reducing total cost of ownership by more than 80% for all types of AI and HPC workloads. Such high-value propositions are expanding NVIDIA’s data center customer base beyond hyperscalers and HPCs to enterprises as more companies use AI and data science.

Check out all the data we have added to our quote pages. Now you can get a valuation snapshot, earnings and revenue estimates, and historical data as well as dividend info. Take a look!