January should be a busy month, with many AI companies reporting or getting ready to report full-year 2019 results and providing an early glimpse into 2020. The beginning of a new year -- and more importantly a new decade, one in which global spending on AI is expected to surge by tens of billions of dollars a year -- is thus time to review the portfolio and make some new additions. Where many articles (rightly) focus a lot of attention on artificial intelligence software development, I instead want to take some time to highlight the hardware that makes it all possible.  

After all, many tech hardware companies were stuck in a rut in 2019 as data center constructors took a breather and the trade war between the U.S. and China took a toll. It appears that slump could be turning a corner in the new year, though. Thus, I think NVIDIA (NVDA -2.55%), Arista Networks (ANET -2.15%), and Micron Technology (MU 1.63%) are in for further rallies in 2020 and should be on AI investor watch lists.

A brain illustrated with electrical connections hovering above a tablet, signifying artificial intelligence.

Image source: Getty Images.

Graphics technology is rocket fuel for AI

While no conversation about AI can start and end with one company, if I had to pick only one, NVIDIA would be it.

It's easy to get hung up on the fact that more than half of NVIDIA's sales are still derived from its graphics processing units (GPUs) for the gaming industry, but video games have served and continue to serve as a valuable basis for the chip designer's future. Graphics rendering computing is finding wide-ranging uses outside of video games, AI being one of them. As a result, NVIDIA has opened the door to new markets the last few years, including self-driving cars and robotics, healthcare imaging, and retail operations management. Even the U.S. Postal Service has become a customer.

Yet NVIDIA is as much a software company as it is a hardware company. It packages its GPUs and other devices with free software for its customers, making it more of a technology platform provider to help organizations jump-start their use of AI. Thus, the company is at the forefront of the movement, powering new breakthroughs like conversational language AI and deep-learning recommendation systems.  

Granted, NVIDIA is no cheap stock, especially for a large cap company currently valued at $142 billion. Forward price to earnings is at 32.3, including expectations that sales will continue to recover from the cyclical dip in sales last year. However, I for one wouldn't be surprised if NVIDIA reclaims and surpasses its all-time high stock price set in late 2018. Shares rallied 74% in 2019, and if revenue continues to trend higher, NVIDIA's share price should follow suit in 2020 and beyond. Ahead of the company's fourth quarter report -- due out in mid-February -- the stock is on my watch list.  

Transferring massive data from point A to point B

According to network hardware giant Cisco, the amount of data traveling through the worldwide web should continue to grow at an average rate of 26% per year through 2022. There are a lot of reasons for the boom in digital data, including the rise of streaming video (video uses way more data than a typical static data file), rising numbers of internet users around the globe, and hundreds of millions of new network-connected devices coming online every year (a movement dubbed the "Internet of Things"). AI is also a data-intensive process and is another reason for the surge.  

Just as roads and freeways need to expand to accommodate more traffic, so do the data centers and spines connecting them to people and machines. Open architecture network hardware and software provider Arista Networks has experienced huge double-digit growth the last few years as a result, and that trend should continue in the years ahead. However, a few headwinds have persisted and knocked the share price around substantially in the last year. Specifically, big names in AI like Microsoft (MSFT 0.55%), Alphabet's (GOOGL 0.49%) (GOOG 0.46%) Google, and Facebook (META -1.64%) are expected to slow down their data center construction in 2020. Arista Networks also pushed back expected deployment of its 400G network technology (think of it like a new freeway upgrade with more lanes) to 2021. 

Nevertheless, in spite of the negative press, Arista still expects to grow in 2020, albeit at a slower rate. Providing hardware and services to the largest public data center operators is but one outlet for the company. Its new Cognitive Campus Edge platform goes after smaller companies looking to build their own networks for private use, especially those looking to make use of AI and who need to keep data private and sectioned off from servers dedicated for public use.  

Thus, with Arista down and out (trailing and forward price to earnings are at 21.4 and 22.5, the lowest they've ever been for the growth stock) ahead of its Q4 report due at the end of January or early February, I'm eyeing another purchase for the long haul.  

Information needs to go somewhere

In addition to networking hardware, the boom in digital data being driven by AI will also create new demand for storage space. After all, once an AI system has been created, it needs to be stored somewhere. Storage takes place in the central data center the AI system was trained and created in, as well as at the location that the system is ultimately deployed to -- like in an "edge" computing application, such as a retailer using AI and smart cameras to monitor and manage a store. 

AI systems are massive, so digital memory demand will ultimately increase in time as network and internet demand also rise. That bodes well for Micron, a leader in manufacturing memory chips and developing new memory technology. In fact, new uses for memory chips like data center-based AI are helping create a floor in chip pricing for the company, and helping it weather the current hardware slump far better than in times past. Though sales were down 35% in Micron's fiscal 2020 first quarter and down 23% during fiscal 2019, the company has remained profitable -- something it has not managed to do in previous chip demand down cycles.

But why Micron now? During the last earnings release, management was upbeat that sales will turn a corner next year, with news that it received licenses from the U.S. government to once again begin doing business with Chinese tech giant Huawei. Micron rebounded 68% in 2019 on anticipation it would eventually return to growth, and those gains should continue if the memory chip maker delivers. With forward price to earnings at just 10.0, Micron is also on my top watch list to kick off 2020.