Given the pivotal role of data and computational prowess in driving innovative competitiveness during the digital epoch, this study expands the scrutiny of environmental impacts stemming from digital technology. It investigates two distinct strata of national-level computing infrastructure (CI)—namely, the National Supercomputing Centers (NSCs) and the National Big Data Centers (NBDCs)—focusing on their influence on urban carbon emissions. Employing the synthetic control method, the research unveils that despite the heightened energy consumption at both the NSCs and the NBDCs (scale effect), these computational infrastructures exhibit disparate effects on urban carbon emissions. This incongruity is principally explicated as follows: the establishment of NSCs amplifies regional carbon emissions, while the carbon augmentation effects of the NBDCs are negligible. Furthermore, through examinations of technological effects and compositional effects, it is ascertained that the CI has not significantly improved the structure of local economic sectors, and its triggering effect on green innovation is only evident in NSCs. These discerning findings elucidate that high energy consumption is a major driver of carbon-intensive outcomes in supercomputing and data centers. Nonetheless, it is imperative to underscore that under favorable conditions, computing infrastructure still possesses the potential to significantly alleviate the adverse environmental “side effects”.