Article by Melissa Croxford
Over the past few months, artificial intelligence (AI) has been making headlines since the creation of ChatGPT which was developed by OpenAI. This AI chatbot led to the widespread mainstream usage of AI and has created a broader discussion on how AI can best facilitate business structures. Following news that chipmaker NVIDIA hit a market capitalisation value of $1 trillion – joining the likes of tech giants Apple, Microsoft, Amazon, and Google’s parent company Alphabet – the financial potential of AI and companies associated with it have been revealed.
NVIDIA began in 1993, focusing on 3D graphics in the gaming and multimedia markets. In this endeavour, it invented a chip called a graphics focusing unit (GPU) in 1999. This invention was key to its recent success as the GPUs facilitated the creation of ChatGPT because the chatbot was trained using over 10,000 NVIDIA’s GPUs. Thus, the success of ChatGPT largely rides on chips manufactured by NVIDIA. This success has led to other companies being in the headlines such as Metaphysic and their deepfakes also using NVIDIA’s GPUs to create their photorealistic videos of celebrities. Seemingly, to be successful in the AI industry NVIDIA’s GPUs seems to be the key with Elon Musk reportedly securing the GPUs for his AI start-up, called X.AI, indicating that this chip is regarded as crucial to the triumph of AI.
The road to this $1 trillion market cap was not solely driven with AI as the destination. Rather, NVIDIA’s focus on gaming helped enable the company to revolutionise the tech industry by setting the foundations of its success with its creation of GPUs for gaming. In 2006, the CEO Jensen Huang decided to invest NVIDIA’s resources in creating a tool to make GPUs programmable. This decision spearheaded the company’s new directional focus as not being solely based on gaming as the chips being programmable had little impact on computer game players but it enabled researchers to perform high performance computing on consumer hardware. Eventually, this led to a breakthrough in modern AI with AlexNet which classifies images. Its model was trained using GPUs, paving the way for the likes of OpenAi to create ChatGPT. Therefore, although NVIDIA’s involvement with AI has seemingly only now come to the forefront of news headlines thanks to the success of ChatGPT, NVIDIA streamlined the production and evolution of AI at several key points.
That is not to say that NVIDIA has abandoned its initial focus on gaming as it recently introduced a new kind of cloud gaming service titled GeForce. Thus, the company has formulated a business model which focuses on different profitable areas of tech. Furthermore, NVIDIA announced at its 2023 Computex keynotes that it plans to showcase its Avatar Cloud Engine (ACE) for games. This conveys a concerted effort by the company to keep both focuses of the company alive as well as indicating the possibility of a fusion of AI and gaming. It is no wonder that NVIDIA has estimated their expected sales reaching $11 billion by the end of July which is more than 50% ahead of the $7.2 billion analysts had predicted.
That is not to say that NVIDIA is alone in the AI-related markets: there are in fact several competitors vying for their top spot. The primary competition includes AMD and Intel, who are both better known for making central processing units (CPUs) but also make dedicated GPUs for AI applications. Whilst neither company is quite at the same level as NVIDIA yet, it is interesting that they have garnered some success by replicating the winning chip model by NVIDIA. What they do with it in future and how they differentiate it from NVIDIA will be interesting to see and compare to how NVIDIA moves forward following the mainstream recognition of its success.
Moreover, several big tech companies, such as Microsoft and Meta, have deployed their own chip projects. Companies like Google have their tensor processing units (TPUs) to aid in search results but also for certain machine-learning tasks. This effect of several of these companies engineering their own chip units demonstrates that AI technology is now firmly cemented in the tech sphere and indicates that there will likely be more success by AI-related companies like NVIDIA.
This success has passed on to external manufacturers as Foxconn, Apple’s largest manufacturer, is expecting demand for servers needed to run ChatGPT-like services to double this year, and with that, so will demand for the products it helps to manufacture. Manufacturing servers for branded vendors such as Dell and other products, Foxconn’s interest in servers running AI chatbots is vast. Foxconn’s revenue last year was $215 billion with a sixth coming from manufacturing servers. Of that server business, 20% was in AI. Indeed, Foxconn has developed an interest in the success of AI because of its heavy involvement in platforms that help these chatbots to operate. The boom of AI seems to be generating financial success in other less directly AI-affiliated companies. Thus, the success of AI has a somewhat trickle-down effect on other related industries.
Translating into companies with recent AI developments, the AI boom has aided Palantir to have its share price more than doubled in the space of 5 months. The company, renowned for its mystique, has experienced this boom thanks to a new AI platform available that focuses on more real-world uses of the technology. Palantir is known for being the tech-world’s version of a management consultancy. The company’s AI platform will, for example, help to identify an enemy tank and suggest ways to target it, something it claims has been used by Ukraine forces. This demonstrates that AI can be used by existing companies to help branch out their businesses and become more profitable.
The NVIDIA market cap of $1 trillion has encapsulated the potential economic success of being involved in the production and use of AI. Whilst it seems various companies are vying for similar success to NVIDIA using AI, the vast uses of AI and the potential for those involved throughout the manufacturing process – including external manufacturers – demonstrates that the market for success in AI is broad enough to encompass all sorts of companies.