Delay in DeepSeek R2 Launch: Is Nvidia the Culprit?

DeepSeek R2, initially expected to launch in May, is delayed due to hardware limitations. Following the viral success of DeepSeek R1 in early 2025, there were high hopes for R2, which aimed to offer major enhancements at lower costs compared to competitors like ChatGPT. Despite innovations that allowed R1 to rival ChatGPT’s capabilities using various AI chips, DeepSeek currently faces challenges acquiring advanced Nvidia chips needed for R2’s development, exacerbated by the Trump administration’s ban on sales to China. CEO Liang Wenfeng is reportedly dissatisfied with the model’s performance, and engineers are awaiting his approval for release. Furthermore, rising demand for existing Nvidia chips among DeepSeek’s cloud customers poses risks to user experience. As the launch remains uncertain, US AI firms benefit from the delay, giving them time to promote their products and optimize costs without immediate competition from DeepSeek. Rumors suggest R2 may offer better coding abilities and multilingual support.

Source link