The artificial intelligence (AI) market -- which includes computer vision, speech recognition, data analytics, and machine learning technologies -- is often considered the next frontier of computing. Research firm Markets and Markets estimates that the market will grow from $240 million in 2014 to $5.05 billion by 2020, fueled by rising demand for AI tech in the media, advertising, retail, finance, and healthcare sectors.
In a previous article, I highlighted 10 fascinating facts about the AI market. Today, I'll narrow the field down to three specific companies that could benefit from its growth -- IBM (IBM -0.85%), NVIDIA (NVDA -1.58%), and Amazon (AMZN -3.31%).
Many investors know IBM as a slow-growth tech company that posted 17 straight quarters of revenue declines. Slow enterprise spending, competition from nimbler rivals, and a strong U.S. dollar have all taken their toll on Big Blue's core business services, tech services, and hardware segments.
But amid all those declines, revenue at IBM's higher-growth "strategic imperatives" (cloud, AI, analytics, mobile, social, and security) rose 12% annually to $8.3 billion last quarter and accounted for 38% of its trailing-12-month sales. Within that higher-growth unit, revenue from cognitive solutions rose 3.5% to $4.7 billion.
The centerpiece of that unit is Watson, an AI platform that processes huge amounts of data for businesses and organizations. Two years ago, IBM set aside $1 billion to create a new business unit for Watson and let it analyze large amounts of data. The platform previously processed huge databases of health data to help doctors, invented its own food recipes, and is now predicting the weather and learning cybersecurity skills to assist IT professionals. The platform also powers AI concierges that assist customers for retailers like Macy's and 1-800-Flowers.
IBM has refused to disclose how much revenue Watson generates, but the market for "predictive analytics" (which helps predict outcomes based on accumulated data) is growing. Research firm Gartner estimates that predictive analytics will capture 40% of net new investments in business intelligence and analytics by 2020.
One of IBM's biggest partners in cloud computing and machine learning is NVIDIA (NVDA -1.58%). IBM installs NVIDIA's Tesla GPUs alongside its Power CPUs in cloud-based enterprise data center systems. Last November, IBM claimed that pairing a Tesla K80 with its Power CPUs enables Watson to answer questions 1.7 times faster than previous systems.
NVIDIA's high-end GPUs provide IBM's Power CPUs with that boost because GPUs are generally better at handling machine learning operations than stand-alone CPUs are. NVIDIA claims that the Tesla K80 can run key science applications "two to five times" faster than Intel's Xeon Phi 7120 CPU. That's why Facebook -- which has been expanding into computer vision, analytics, and machine learning -- installed NVIDIA's Tesla M40 GPUs in its Big Sur computing platform last year.
NVIDIA's data center revenue, which includes its Tesla and GRID platforms, rose 63% annually last quarter and accounted for 11% of its top line. That business will likely keep growing -- along with demand for high-powered machine learning GPUs.
Amazon's AWS (Amazon Web Services) is the biggest cloud platform in the world, with an annual run rate of about $10 billion. It also generated over half of Amazon's operating profits last quarter, despite only bringing in 9% of its sales.
AWS evolved out of the backbone of Amazon's e-commerce site. That backbone had so much computing power that Amazon started hosting other major sites, like Netflix, NASA, and the CDC, on its servers. Last year, it launched Amazon Machine Learning, a service that synchronizes to AWS' online storage service S3, its data warehousing solution Redshift, and its Relational Database Service (RDS).
Like Watson, Amazon Machine Learning refines the data on its AWS servers into actionable business predictions through analytics. For example, it can help a retailer predict how gender, income, and profession impact purchases of certain products. These predictions, which Amazon sells in big batches, could prove invaluable to marketers. Amazon also uses predictive analytics on its e-commerce site to recommend products to customers based on their past purchases.
The key takeaway
Facebook and Alphabet are both major players in AI, but their AI efforts mainly enhance their core advertising businesses instead of acting as stand-alone sources of revenue. Investors who want more direct exposure to the AI market should take a closer look at IBM's Watson, NVIDIA's Tesla GPUs, and Amazon's Machine Learning platform instead.