As technology companies are using NVIDIA’s GPUs to meet their growing AI needs, the market may face a shortage of GPU supplies, especially to meet the needs of gamers first.
In addition to integrating ChatGPT, Microsoft’s Bing search engine is seeing an increase in demand for other generative AI tools. These generative AIs significantly rely on the processing power of artificial intelligence, which is where NVIDIA GPUs have an advantage.
NVIDIA GPU is frequently utilized in the field of deep learning since it offers Compute Unified Device Architecture, and Unified Computing Architecture, where AI models use GPU for computing. According to a report on the FierceElectronics website, 10,000 NVIDIA GPUs were used to train ChatGPT at first.
Since it came to the attention of the public, the system has become overburdened and unable to handle the demands of numerous users. To do this, OpenAI introduced the ChatGPT Plus subscription service, which not only enables users to use ChatGPT before others at peak times but also has a quicker response time and can prioritize the use of the newest features.
As recently observed by industry experts, the ChatGPT model is currently being trained on around 25,000 NVIDIA GPUs, up from the 10,000 NVIDIA GPUs used in initial testing.
Microsoft and Google have announced incorporating ChatGPT-like LLM into their search engines. To integrate such an AI model into search results, using Google as an example, it is projected that 512,820 A100 HGX servers and 4,102,568 A100 GPUs are needed. It is predicted that the capital investment will reach up to $100 billion in server and network costs alone.
Although the growth of generative AI is fantastic for NVIDIA, it may not be so beneficial for consumers, especially gamers. NVIDIA might put the needs of technology providers of generative AI ahead of those of gamers.
Due to the Lunar New Year, there is a shortage of gaming graphics cards during the first quarter of this year. Even while the inventory issue remains, it could cause supply issues given the limited availability of premium GPUs.
To meet the rising demand for generative AI, high-end GPUs, which offer the AI capabilities that make NVIDIA profitable – might be in short supply. How NVIDIA will respond to the enormous demand in the AI area is yet to be seen. By February 22, the company is anticipated to release its financial report for the fourth quarter of fiscal 2023, during which it is anticipated to provide further details regarding its operating expectations going forward.
