With the evolution of edge devices and the advancement of AI, the significance of accelerator cards has been increasing. There is a growing demand for processing large volumes of data from edge devices and conducting temporary analyses. To meet these needs, distributing tasks using accelerator cards has become crucial. This helps reduce CPU load and focuses not only on data processing but also on storage and security. The downloadable document provided here offers a comprehensive overview of the challenges and technological trends faced by accelerator card designers, along with tailored proposals for each unit. We hope this information will be valuable to everyone involved in accelerator card design, and we encourage you to download and review the content.