DeepSeek built a competing large language model (LLM) to OpenAI's ChatGPT, but claims that it trained the model with old Nvidia processors that aren't widely used anymore. This news has shellshocked technologists and analysts across Wall Street because if DeepSeek's methods of building highly capable AI can be done using legacy hardware,

DeepSeek stunned the tech world with the release of its R1 "reasoning" model, matching or exceeding OpenAI's reasoning model for a fraction of the cost.

Because Nvidia is still growing its revenue rapidly, using trailing earning metrics can give a false impression of where the company is headed. However, they're also useful in this case because its future had a wrench thrown into it thanks to DeepSeek's breakthrough.