The trillion-dollar semiconductor manufacturer has completed more than twenty deals, utilizing both its Nventures investment division and corporate development team. Companies within its portfolio emphasize the added benefits of such agreements, including access to research teams and even the renowned CEO of the company.
By Richard Nieva, Forbes Staff
When Terray Therapeutics, an artificial intelligence biotechnology company, initially trained its AI model Coati in the early part of last year, CEO Jacob Berlin found the technology to be “fairly satisfactory.” While functional, Berlin believed there was room for improvement.
However, following an investment from Nvidia, the trillion-dollar semiconductor giant, specializing in small-molecule drug discovery, Terray completely retrained the model. With enhanced computing resources and Nvidia’s engineering expertise, Berlin noted a significant enhancement in performance. “We witnessed a substantial improvement,” Berlin shared with Forbes. “Our collaboration with Nvidia was essential in achieving this progress.”
The transformation of the AI model exemplifies the intangible advantages associated with having the world’s most valuable semiconductor company as a partner. This holds true for over two dozen companies that Nvidia has invested in over the past couple of years, facilitated by both its venture capital arm Nventures and its broader corporate development team.
For Terray, the crucial factor was the accessibility to Nvidia’s researchers and engineers, renowned for maximizing the potential of the company’s highly sought-after GPUs (graphics processing units) utilized in extensive AI development projects. In certain instances, this access translates to prompt responses from CEO Jensen Huang, an increasingly influential figure in the technology sector, who oversees Nvidia’s investment approval process. For many startups, it signifies gaining validation from one of the key players in the ongoing AI frenzy.
Sid Siddeek, corporate vice president and head of Nventures, stated to Forbes, “Our investments are comprehensive—not just in terms of capital but also in expertise, technology, advantages, and various other aspects.” He emphasized that one of the primary objectives of these investments is to “generate a profitable return for Nvidia, akin to any financial investor.”
Many major tech companies operate their venture capital arms, including Alphabet, Microsoft, and Salesforce. However, Nvidia, a relatively recent entrant in this domain, offers unique benefits beyond those of a traditional venture capital firm, as noted by portfolio companies. The establishment of the unit occurred in early 2022, preceding the emergence of OpenAI’s ChatGPT, predominantly trained on Nvidia GPUs, which triggered a surge in AI adoption across Silicon Valley. Since then, Nvidia has been actively investing in prominent AI unicorns such as Cohere, Hugging Face, and Inflection.
Siddeek remarked, “The rise of gen AI was notable. We transitioned from ground zero to a remarkable point. It’s an exponential growth from zero.” By December, the company had announced 14 investments since the commencement of 2023, spanning various sectors like health, enterprise, and logistics. Reflecting the ongoing expansion, Nventures’ website was launched just last month.
While Nventures has already finalized several investments in the new year, as disclosed by Siddeek to Forbes, detailed specifics were not provided.
Forbes conversed with seven portfolio companies of Nvidia, who opted not to disclose the precise terms of their agreements with the semiconductor giant. Some mentioned receiving improved access to Nvidia’s GPUs, currently in high demand within Silicon Valley, without divulging specific details. Others indicated that they did not receive any preferential treatment.
Jorge Torres, CEO of MindsDB, expressed, “Describing it as a VIP pass might be overstating it, but the relationship does provide certain advantages compared to not having it.”
Nventures clarified to Forbes that they do not offer prioritized access to GPUs. As startups and tech behemoths engage in a competitive race to train the most advanced AI models, GPUs have emerged as a new form of currency. However, during the negotiation phase, Nvidia’s portfolio companies did inquire about this aspect, as mentioned by Kanjun Qiu, CEO of Imbue AI, a startup focused on developing AI capable of human-like reasoning. Qiu humorously mentioned, “We understood that preferential access was unlikely. Even though we asked, ‘…will we get preferential access?’” Nvidia and several founders of portfolio companies affirmed that there were no obligations or directives for startups to allocate any portion of the capital towards Nvidia chips or other products.
Both Nventures and the corporate development team operate based on Nvidia’s financial resources. Presently, it is an opportune time for such investments. Nvidia dominates the GPU market, accounting for over 70% of AI chip sales, according to research by Omdia. In the initial three quarters of the previous year, the company amassed \(38.82 billion in revenue (with the announcement of fourth-quarter earnings pending). Notably, sales surged by more than 200% in the third quarter compared to the previous year. Consequently, Nvidia’s market capitalization skyrocketed to \)1.8 trillion this week, positioning it as the fourth most valuable company globally, surpassing Alphabet, Amazon, and Meta.
Got a tip about Nvidia? Reach out securely to reporter Richard Nieva on Signal at 510.589.4118 or [email protected].
Another compelling aspect of an investment from Nvidia is the accessibility to Jensen Huang, the CEO known for his signature leather jacket. While the extent of interaction with the CEO varies among portfolio companies, Huang played a role in the initial engagement process for companies like Imbue and Utilidata, a firm dedicated to revolutionizing the energy grid through AI. Qiu, the CEO of Imbue, affirmed that Huang has been approachable post-investment. She recounted an instance when she promptly received a response from Huang after emailing him a query regarding executive compensation practices. “He is quite reachable. I try not to disturb him excessively, but he is certainly approachable.”
Conversely, some founders of portfolio companies admitted to never meeting Huang. Siddeek noted, “It’s essential to remember that Jensen has numerous responsibilities. This is quite evident.” However, he underscored that Huang leads the investment committee responsible for approving every deal. Nvidia refrained from commenting when questioned about Huang’s criteria for engaging with founders during the deal negotiation process.
Apart from Huang, founders highlighted the prompt responses they receive from Nvidia as portfolio companies. “It is evident that within Nvidia, there is a rapid response to requests for assistance from the Nventures team,” stated Bob Hall, CTO of Outrider, a company specializing in autonomous freight and logistics. “Thus, the access they provide and the significance attributed to Nventures internally at Nvidia have been immensely beneficial for us.”
While Nvidia currently leads the tech industry, rapid shifts in Silicon Valley are commonplace (as evidenced by the experiences of Yahoo, MySpace, or Huang’s former employer, AMD). As competitors strive to develop viable alternatives to Nvidia’s chips, the company might view its startup investments as a strategy to diversify its business in the event that Nvidia’s GPUs lose their current market dominance. Three founders of Nventures portfolio companies shared their belief with Forbes that this could be a potential scenario.
“It may prove challenging to displace Nvidia in the upcoming generations, but undoubtedly, there are exceptionally talented individuals who will attempt to do so,” mentioned one founder, requesting anonymity to candidly discuss Nvidia.
Siddeek acknowledged that from an investment perspective, the company is forward-thinking. “We seek founders with a bold vision,” he stated. “Individuals who envision the world not as it is today but as it could be in five or ten years.”
Nvidia has a commendable track record in this regard. Established three decades ago with a primary focus on enhancing video game graphics through its processors, Nvidia adapted its business direction in recent years towards crypto mining. Most notably, GPUs have emerged as the primary engine for AI training.
The breakthrough occurred in 2009 when a group of Stanford professors, including Andrew Ng, who later co-founded the Google Brain research lab, recognized the efficient computational structure of GPUs for accelerating machine learning training. Nvidia’s success can be attributed to being at the right place at the right time, coupled with Huang’s strategic decisions to capitalize on the impending AI revolution.
Terray’s Berlin remarked, “Nvidia has a remarkable track record of identifying verticals poised for significant growth and demand for their chips and software.” “They aim to cultivate another multi-billion dollar vertical within the company.”
Alex Konrad, Iain Martin, and Kenrick Cai contributed to the reporting.
