Press reports, citing inside sources, indicate that DeepSeek has not yet set a new release date for its R2 model, which was originally scheduled to launch this past May.
The decision to postpone the launch was reportedly driven by the company's CEO, Liang Wenfeng, who expressed dissatisfaction with the model's current performance level.
In response, the company's engineers have been working for several months to enhance R2's capabilities to meet Liang's ambitious standards.
The new model was expected to be a significant leap forward from its successful predecessor, R1, particularly in its coding abilities and understanding of languages other than English.
The Chip Crisis... Another Hurdle for the Anticipated Model
Separately, the company faces an equally significant challenge: a severe shortage of the NVIDIA server chips essential for running advanced AI models.
Reports have pointed out that most cloud computing clients currently using the R1 model rely on NVIDIA's H20 chips, a component specifically designed for the Chinese market.
However, new restrictions imposed by the U.S. administration last April have prevented even these chips from reaching China.
Employees at major Chinese cloud computing firms have warned that a sudden, high demand for the R2 model could overwhelm their infrastructure due to their limited chip supply.
The delay follows the unexpected success of the R1 model upon its release. The model caught the AI sector by surprise with its efficient performance relative to its low cost.
R1 demonstrated that Chinese companies can still compete fiercely despite the technological restrictions placed upon them.
In an effort to prepare for the future, DeepSeek is coordinating with cloud companies in China, providing them with the necessary technical specifications to plan for hosting and distributing the new model once its launch is approved.
But the question of when the R2 model will finally see the light of day remains, caught at the intersection of high technical ambitions and a complex political and economic reality.
Summary
Chinese company DeepSeek is postponing the launch of its anticipated R2 model. The decision stems from two main challenges: the CEO's dissatisfaction with the model's current performance and the significant hurdles posed by an advanced electronics chip shortage caused by U.S. restrictions.
