Nvidia’s ChatGPT-powered profits are predicted to have a greater impact on gamers than crypto ever did. The rise of generative AI, produced by large language models such as Midjourney and ChatGPT, poses a threat to the availability of gaming GPUs. This could lead to a shortage of gaming graphics cards, similar to what occurred during the peak of the crypto craze. The value of crypto has been hotly debated, with some suggesting that it has very little real-world application. In contrast, generative AI has many practical uses, including assisting in film production, creating music and drafting documents in various industries.
However, the use of AI-generated media has been criticised as dehumanizing. The Writers Guild of America is currently on strike, with one of their primary concerns being the possibility of studios using AI to generate new scripts instead of using actual writers. Unfortunately, if studios or media companies could make money from AI-generated media, even if it were of inferior quality, they would do so as it cuts labour costs and increases profits.
The AI technology behind generative AI is based on neural networks that require specialised tensor cores within graphics cards to function effectively. Nvidia’s GPUs are currently the most effective for carrying out matrix multiplication essential for machine learning and to produce useful generative content. Other competing AI hardware from Intel and AMD is not as mature as Nvidia, making their GPUs less effective for training neural networks. As more companies pivot away from Web3 towards generative AI, the demand for AI hardware will only grow.
Nvidia’s tensor cores are also the driving force behind its record profits last quarter, turning the company into a trillion-dollar industry almost overnight. Many players in the industrial sector, such as Google and Microsoft, need large-scale GPU operations to continuously train their AI models. Therefore, there is a lot of demand for Nvidia hardware, and the market is adapting to this new reality.
However, the potential for independent operators to have a place in this new paradigm through distributed computing and processing of training data is also feasible. This would allow independent operators to rent hardware when needed, rather than purchasing it outright, reducing overheads. Nevertheless, there is only so much supply of Nvidia hardware, and sales of their midrange GPUs have been fairly flat among gamers this year, so the market is still adapting to the new reality.
Regardless, distributed computing platforms such as Folding@home demonstrate how easy it is to roll out distributed data processing tools. Twitter users are already bemoaning the lack of available graphics cards even though the AI phenomenon is still in its infancy. It seems gamers may face a new issue soon: scarcity of graphics cards due to AI’s growing demand for visual processing power.
