banabana-san/iStock via Getty Images
banabana-san/iStock via Getty Images
Artificial intelligence ("AI") is briskly revolutionizing numerous industries, and NVIDIA Corporation (NASDAQ:NVDA) is an undeniable leader in developing AI hardware and software. To maintain its competitive edge, Nvidia has strategically invested in several key sectors in the AI space. Whilst it is true that these investments are insignificant in terms of immediate financial returns, they hold the potential to substantially enhance Nvidia's future AI capabilities and bottom line.
Back on the 14th September 2023, it was announced that Nvidia had participated in Databricks’ Series I funding, which raised over $500 million. In the press release, no information was given regarding the size of Nvidia’s investment in Databricks, although, the Form 10-K on page 52 shows an increase in “other,” which would include non-marketable securities, of $680 Million, indicative of a sizable contribution to the $500 million raised, although, this is chump change for this $2.16 Trillion dollar company.
On Wednesday 14th February, Nvidia disclosed in a filing that it held publicly traded stocks in five companies linked to AI: Arm Holdings plc (ARM), SoundHound AI, Inc. (SOUN), Recursion Pharmaceuticals (RXRX), Nano-X Imaging (NNOX) and TuSimple (TSPH). The amount invested in each of these companies also amounted to a paltry sum for a company the size of Nvidia.
Even if the value of these investments were to triple, the overall profit would be microscopic compared to the last quarterly earnings of over $12 Billion. However, it is my contention that, combined, these investments are underappreciated by Nvidia investors, and will, in time, enable Nvidia to remain the leader of the AI pack. Personally, I believe, if I may quote Jnr Choi’s finely crafted lyrics, “the ting gon turn up.” That is to say, I expect Nvidia to continue its rise for the foreseeable future.
In the article, I describe how three of Nvidia’s investments will, potentially, boost the company’s earnings going forward.
Nvidia invested in privately held Databricks, a rapidly growing leader in data analytics with a user-friendly cloud platform. Databricks focuses on open-source tools, attracting a large developer base and propelling them to acquire more customers. This strategy positions them well for the booming big data market.
There has been an impressive surge in the number of customers Databricks has signed up to their offering. In 2021, the company stated that “more than 5,000 organizations worldwide” used Databrick’s platform but in 2023, the company claimed to have more than 10,000 customers.
The partnership with Nvidia will, I believe, help to solidify their place at the forefront of AI innovation.
Obviously, Databricks' software runs on top of computer hardware, and Nvidia is a major player in that space with their graphics processing units (GPUs). By working together, they can optimize Databricks' software to run better on Nvidia GPUs, making both companies' products more attractive. (I cover Databricks in more detail in a recent article I wrote about Snowflake Inc. (SNOW))
Secondly, Databricks is a leader in data analytics, a field that's crucial for developing AI. By having a stake in Databricks, Nvidia gains a partner that can help them develop better AI tools and solutions. This can strengthen Nvidia's position in the overall AI market. During the recent earnings call, Nvidia's Chief Financial Officer Colette Kress stated that the company "also made great progress with our software and services offerings, which reached an annualized revenue run rate of $1 billion in Q4." This is a big deal, because it showcases Nvidia's ambitions to build beyond semiconductors.
In short, this investment has created a win-win situation. Nvidia gets to partner with a leading AI software company, and Databricks gets access to Nvidia's powerful hardware.
Arm Holdings, a British-based chip design giant, has revolutionized mobile computing. They license their energy-efficient designs, powering over 90% of smartphones from Apple, Samsung, and Qualcomm. Their reach extends beyond mobile to IoT devices, servers, and even laptops.
Nvidia, the leader in graphics processing units (GPUs), attempted a record-breaking acquisition of ARM in 2020. This aimed to combine the two chip powerhouses, but concerns over competition and innovation led to a failed deal. However, Nvidia retains a substantial stake in ARM.
Despite this failed merger, collaboration opportunities remain. Both are leaders in AI, and their technologies could work together in areas such as self-driving cars and high-performance computing.
To begin with, there is an opportunity to embed Nvidia’s technology into an ARM design. Nvidia's AI processing cores or specialized accelerators could be licensed and incorporated directly into the ARM CPU design. This would allow ARM processors to handle basic AI tasks on-device without needing a separate Nvidia GPU.
In addition, perhaps there is a way for Nvidia’s software to be integrated into a customer’s solution. Nvidia's AI software frameworks, such as TensorRT, could be optimized to run on ARM CPUs. This would leverage existing ARM processing power for specific AI functions, while still relying on Nvidia's expertise for the software tools.
Both approaches would extend Nvidia's AI technology beyond their GPUs, creating a royalty stream from chip manufacturers licensing the tech for ARM CPUs in various devices.
Recursion Pharmaceuticals is a company that utilizes AI for drug discovery. Recursion's approach leverages machine learning to identify promising drug candidates. Potentially, this could significantly reduce the time it takes to discover new medicinal compounds. Their automated lab facilities churn out a staggering number of experiments, generating massive datasets for analysis. The AI "learns" by predicting potential drugs, synthesizing the most promising ones, and testing them in the lab. These results, positive or negative, continuously refine the model.
Recursion uses massive datasets for drug discovery, which requires immense computing power. Nvidia GPUs excel at processing large datasets efficiently. By collaborating, they can significantly accelerate the drug discovery process.
The statistics the company offers are impressive. Its website claims its automated lab robotics enables it to conduct “up to 2.2 million experiments each week for up to 50 weeks per year,” All these experiments and trials generates an enormous amount of “proprietary high-dimensional data.” The idea is that the machine "learns" through predicting medicinal candidates for a particular malady, synthesizing the most promising ones and testing to see if they are potentially curative. The results of these tests, whether positive or negative, direct the machine’s future direction.
Recursion's vast biological and chemical data provides a valuable training ground for Nvidia's AI models. This data can be used to refine AI algorithms specifically for drug discovery, leading to more accurate predictions and potential breakthroughs.
Nvidia's BioNeMo platform is designed for AI-powered drug discovery. By working with Recursion, Nvidia can showcase BioNeMo's capabilities and potentially attract other pharmaceutical companies to the platform, creating a new revenue stream.
SoundHound, a voice AI company, has a market cap of over $2.5billion-dollars with revenues moving in a very healthy direction. Their stock has soared over 398% in just a year. Founded in 2005 as a music identification rival to Shazam, SoundHound now offers a broader range of AI-powered voice features. These include automated restaurant ordering and "intelligent transcription," which goes beyond speech-to-text by recognizing topics and meaning.
The Nvidia investment and the company's expanding voice AI solutions suggest a promising future, even if the path to profitability remains a mystery.
By leveraging SoundHound’s expertise in complex voice recognition and natural language understanding (NLU). This could be integrated into Nvidia products like gaming consoles, self-driving cars, and smart assistants. Imagine voice commands in games that understand context or highly accurate voice control in a car.
Voice solutions could become customizable since SoundHound's platform, Houndify, allows developers to build custom voice experiences. Nvidia could offer pre-built voice solutions powered by SoundHound for developers using their hardware, making it easier to create voice-enabled features. A significant win-win for both companies.
SoundHound's technology excels in noisy environments. This would be crucial for voice assistants in living rooms or cars, or for voice chat in online games.
Clearly, by integrating SoundHound's capabilities, Nvidia can significantly enhance the user experience of their products with more natural, accurate, and versatile voice interaction.
Whilst there is likely very little prospect of gaining a substantial profit on investments in the companies listed above, there will likely be a significant boost to existing Nvidia Corporation product lines as a result of these investments