<p>With the increasing adoption of secure outsourcing computation, clients with limited resources can offload computationally intensive tasks to more capable servers. Matrix multiplication, a crucial mathematical process, plays a key role in neural networks and demands significant computational power. However, currently designed edge server-based matrix multiplication outsourcing schemes cannot ensure data privacy during task distribution and traceability of edge servers that return incorrect results. To address this problem, we explore how to implement parallel outsourcing using edge servers to accelerate matrix multiplication computation while ensuring data privacy, and propose the first distributed verifiable and traceable utility scheme. Firstly, we introduce a novel approach for dividing and allocating subtasks, ensuring the original matrix input remains hidden. Using the designed novel strategy, we partition the computational task into multiple subtasks and ensure that these subtasks can be outsourced to the edge servers in parallel without disclosing sensitive information. In addition, we propose a verifiable and traceable method for matrix multiplication tasks that use parallel edge servers. This method only requires the client to perform a constant number of multiplication and addition operations, which minimizes the computational load on the client. The correctness, security, verifiability, and traceability of the proposed scheme are evaluated rigorously. Finally, we conduct comprehensive experimental simulations to evaluate the feasibility and effectiveness of the proposed scheme.</p>