How do you build the next NVIDIA?

To build a successful company, you must not only be right about the idea, technology and market, you must also execute

AS AN INVESTOR, I have been fairly good at spotting trends, getting in early on fiber optics, bioinformatics, the internet, ecommerce, cyber, cloud and quantum computing.

What seems to be my best bet so far has been on artificial intelligence. I was certainly not alone in identifying the potential of AI to transform not just computing but our entire world, but the founding of OurCrowd in 2013 coincided with the rise of AI and machine learning startups, so we have grown up together. To date, we have invested more than $380 million in AI startups, creating a portfolio of more than 80 AI firms with 8 exits achieved so far.

I failed, however, to link the growing adoption of AI with the meteoric rise of NVIDIA, which this month briefly overtook Microsoft and Apple to become the world’s most valuable company, propelled by its pivotal position as the world’s major supplier of high-performance chips and software for the fast-growing AI market.

Like many others, I have watched in amazement as NVIDIA’s Nasdaq stock price has increased more than 27,000% from 49 cents in June 2014 to more than $135 in June 2024, kicking myself each time as I failed to invest.

To be fair, even NVIDIA did not see this coming a decade ago. The phrase “artificial intelligence” did not even appear in the company’s 2014 Annual Report, which singled out gaming, data centers and automotive as the main drivers of its future growth. Then researchers at Google, Microsoft, Facebook and university labs hit upon NVIDIA chips as the go-to processors for GPU-accelerated deep learning systems for artificial intelligence. By 2016, NVIDIA reported that the number of companies it was collaborating with had jumped nearly 35X to more than 3,400.

“Deep-learning breakthroughs have sparked the AI revolution,” NVIDIA told its shareholders that year. “Progress is exponential. Adoption is exponential. The impact on the tech industry and society will also be exponential.”

Most investors, including me, did not yet grasp the implication of these prophetic words. By the end of 2016, the company’s share price was still less than $3.

Meanwhile, NVIDIA chips were being snapped up as the only processors capable of handling the huge demands of AI. However, they were originally designed as video gaming chips. While they possessed awesome processing power and speed, they were expensive and had high power demands. Mention of the terms “artificial intelligence” and “AI” first appeared just four times in the company’s annual report in 2015, increasing to 17 mentions in 2016 and 70 in 2017, but it was still not the central focus of the business. The company’s 2017 Annual Report opened with a quote from Forbes declaring that “NVIDIA GeForce has moved from graphics card to gaming platform.” By the end of 2017, despite the rapid growth of AI, the NVIDIA stock price was hovering around $5.

In 2017, OurCrowd led a seed investment in Hailo, a fledgling Israeli startup developing semiconductors capable of performing the data-intensive processing demanded by AI and machine learning on the edge instead of sending the data back and forth to centralized data centers. Hailo was one of the first companies to implement an architecture for a processor chip family that was designed specifically for AI from the ground up – and one of only a handful of developers promising AI processing at the point of need in your car, smartphone, tablet, camera, phone or IoT sensor – instead of relaying everything via the cloud to a remote AI processing farm.

When we began our investment in Hailo, we were excited by the great potential for the technology in this burgeoning market, but it did not even cross our minds that we were investing in the next hot thing, that one day would catapult NVIDIA to the peak of global tech company valuations. We invested in Hailo because it was poised to advance AI at the edge. We believed in the basic idea that it made more sense to process AI at the point where it was needed. For that idea to work, you needed to have proven AI capability, and you needed to show it could be done at the edge with low power and at low cost.

But to build a successful company, you must not only be right about the idea, technology and market, you must also execute. Several elements of that execution are critical.

Since 2017, we have seen major, unexpected breakthroughs in the AI sector, including the rapid development of large language models, the growth of generative AI and the rise of companies like OpenAI which have popularized the technology beyond business use into the global commercial market and popular consciousness.

The transformation of the AI market really does change almost everything, opening up a hitherto unforeseen opportunity for other companies with niche technologies better suited to different specialized sectors of this huge new global market.

We now find ourselves as one of the largest shareholders in a company with a billion-dollar valuation, proven technology and multimillion-dollar funding from major investors that is in pole position to play a pivotal role in the further development of an unprecedented commercial surge underpinning the tsunami driving the world’s most valuable corporation. 

I sit on Hailo’s board, representing OurCrowd’s significant stake in the company. It’s a privilege to observe the contours of this revolution close up. I also feel a major responsibility to represent the best interests of our shareholders, as well as making sure that this company fulfills its potential to help shape this historic moment in global technology.

Having picked this promising investment, I must now work with my Hailo board colleagues to decide what we need to do as a company to execute successfully on a far-sighted strategic plan originally formulated more than seven years ago. MBA courses are littered with the stories of companies with great technology and products that failed to execute at a critical moment, or were overtaken by competitors – as Jerry Seinfeld recently illustrated with epic breakfast battle between Kellogg’s Pop Tarts and Post’s Country Squares.

One key area is personnel. Startup founders and technologists are not always the right people to guide a growing company to commercial success. Hailo’s founders, who were in their twenties when they set up the company and have done a brilliant job building a team of more than 200 people, winning major customers and investors, and bringing Hailo to its unicorn valuation, realized that they need experienced, veteran hands to help take their creation to the next stage. They have appointed Yaron Garmazi, a seasoned business executive with proven successes, as Hailo’s new CFO and COO, adding his deep experience of global and financial and operational management at the highest level to the company as it charts its future course.

Another factor in the success of a company is ensuring that it raises enough capital to finance its growth. Here again, Hailo has done the right thing, most recently securing a $120 million financing round at a $1.2 billion valuation, raising its total financing to date to more than $340 million.

Hailo is active in several sectors, including automotive, cybersecurity and other applications. The company will require the correct strategic guidance to choose which market to address with its technology.

We also need to continue to introduce the company to the best possible partners – commercial, industrial, technological, investment – in multiple ecosystems around the world and make sure it is working with the biggest players.

We need to help the company build its manufacturing capacity, especially in a market where massive demand can come out of nowhere, and ultra fast. It will need a supply chain that is not only robust and reliable but also nimble and responsive. Semiconductors usually require long lead times but Hailo’s architecture is much more flexible and adaptive. The company may also need to grow by acquiring new business units and capacity though M&A transactions.

I’ve had the privilege of helping several startups grow to billion-dollar valuations. As hard as that is, it will be even harder to take a company from a billion dollars to a hundred billion, or maybe a trillion, some day. At the end of 2014, NVIDIA’s market cap was $10.9 billion – today it is more than $3 trillion.  Now that I find myself an investor and a board member of a company in the world’s fastest-growing technology sector, I’m looking forward to that challenge. I believe Hailo has the right technology, team and timing to emulate at least some of the success of NVIDIA.

If you have any advice or thoughts, I’d be delighted to hear them.

About ‘Investors on the Frontlines’

I’m the CEO and Founder of OurCrowd, the global equity investment platform that gives individual accredited investors access to private market startup deals alongside top-tier VCs. If you are an investor, private family office or financial advisor, subscribe here for my biweekly commentary or follow me on Twitter. I welcome your comments in the response section below.