Data weaving is a technique that splits input data into multiple parts and processes them in parallel on different computing nodes. In traditional distributed computing, data is usually distributed to different computing nodes for independent processing, but this approach may cause data transmission overhead and communication delays. This article weaves input data and allocates computing tasks to different computing nodes for parallel processing, thereby reducing data transmission and communication overhead. In order to evaluate the performance of distributed computing methods based on data weaving, this article designs a series of experiments and compares the performance differences between this method and traditional distributed computing methods. Experimental results show that the method based on data weaving can significantly reduce data transmission and communication overhead, improve computing efficiency, and have high processing consistency.