Scientific and technological innovation is a core factor in developing new quality productive forces. As an important platform for enhancing the independent innovation efficiency, new infrastructure promotes scientific and technological innovation and industrial change by innovatively allocating production factors and facilitating the flow of R&D capital, labor, and data factors. Based on panel data of 31 provincial-level regions in China from 2012 to 2022, this paper measures the level of new infrastructure, R&D efficiency, and transformation efficiency. A two-way fixed-effects model of time and individuals is used to empirically test the nonlinear impact and mechanism of new infrastructure on innovation efficiency. The results show that the impact of new infrastructure on R&D efficiency presents an inverted U-shaped relationship, while the impact on transformation efficiency is not significant. This result holds after a series of robustness tests. The mechanism tests reveal that capital mismatch does not play a significant role in the process of new infrastructure affecting the independent innovation efficiency, while labor mismatch and market allocation of data factors can play a role. The heterogeneity test indicates that, from the perspectives of three dimensions of new infrastructure, there is a significant U-shaped relationship between information infrastructure and independent innovation R&D efficiency/transformation efficiency, while the impact of integrated infrastructure and innovation infrastructure on the independent innovation efficiency is not significant. From a regional perspective, this paper finds that the new infrastructure has a significant inverted U-shaped impact on independent innovation R&D efficiency in the eastern region, which has a significant inverted U-shaped impact on transformation efficiency in the central and western regions. Based on the above findings, it is proposed to construct a policy combination of “short-term effect and long-term empowerment”, to reconstruct the mechanism of innovation resource allocation through the market-oriented reform of data elements, and to address the spatial imbalance between new infrastructure and innovation efficiency with regional differentiation strategies.