This document summarizes a research paper on implementing a discrete cosine transform (DCT) core using distributed arithmetic with an error-compensated adder tree to improve accuracy and throughput. It introduces distributed arithmetic as an alternative to multipliers to reduce hardware costs in DCT designs. Previous work achieved higher speeds by parallelizing shifting and addition but incurred large truncation errors. The proposed error-compensated adder tree operates shifting and addition in parallel while compensating for truncation errors. It allows using 9-bit precision instead of 12-bit to meet accuracy requirements, reducing hardware costs. The design achieves a throughput of 1 billion pixels per second with a gate count of 22,200 gates.