AI chip startups are seeing a surge in demand for inference chips, largely driven by the adoption of DeepSeek's innovative open-source model.

DeepSeek's advancements are expected to accelerate the transition from AI training to inference, which is the application of AI models to real-world tasks.

Industry experts assert that smaller open models can rival larger proprietary ones in effectiveness while significantly reducing costs, thereby enhancing the age of inference.

Feldman emphasized that the rise of open-source models will prevent any single company from dominating the AI market, challenging traditional hardware and software barriers.

As AI demand continues to grow, smaller companies are poised to compete more effectively, especially as larger players like Nvidia may struggle to meet the increasing demand.

DeepSeek's recent model has notably impacted Nvidia, leading to a substantial reduction in its market capitalization and prompting smaller AI firms to explore new growth opportunities.

Many AI companies view DeepSeek's emergence positively, as it allows them to shift away from costly models like those from OpenAI towards more affordable open-source alternatives.

In response to DeepSeek's advancements, companies are reallocating their budgets from training clusters to inference clusters, reflecting a strategic shift in AI resource allocation.

Analysts believe that DeepSeek's engineering innovations will not only lower inference costs but also improve overall training expenses, fostering broader AI adoption.

The demand for inference chips is rising because they require less powerful hardware, creating new opportunities for startups in the AI chip sector.

Inference, which involves applying AI for predictions, contrasts with the more resource-intensive training phase, highlighting the efficiency of newer models.

Wedbush analysts predict that the increasing usage of AI across various sectors will drive significant demand for AI technologies in the near future.