在半導體產業中，目前皆以電子設計自動化工具軟體來設計積體電路，其基礎設施（Infrastructure）包含電子設計自動化工具軟體運算節點、網路及儲存設備系統。由於製程不斷的演進、縮小，晶片設計的資料量幾乎成等比級數的方式增加，為了讓晶片設計加速、生產時間縮短，整個電子設計自動化工具軟體基礎設施的整合就顯的非常重要。 本研究的目的藉由個案研究的方式，找尋在運算節點、網路及儲存設備系統可調整之參數及設定，利用參數及設定之調整，來提昇整體之效能。個案公司R公司是一家積體電路設計公司，以積體電路產品之研發與設計為主，產品線橫跨通訊網路、電腦週邊、多媒體等。利用深入訪談研究方法，來瞭解個案公司R公司電子設計自動化工具軟體環境及架構，以及電子設計自動化工具軟體與儲存設備系統的整體架構使用狀況。並以Ｒ公司之專案，以系統化、結構化的方式實測，找出最佳的儲存設備系統架構，以及儲存設備系統的快取（Cache）模式。 在R公司測試的結果發現，NetApp儲存設備系統Flash Pool的快取效能優於Flash Cache快取，而且每一個NetApp儲存設備系統控制器規劃兩個資料「聚合」及四個以上的「卷」時，R公司專案的執行時間有明顯的減少。因此，以Flash Pool的快取搭配上述的儲存設備系統架構，可以使得電子設計自動化專案整體的效能提昇。本研究之研究結果，可以作為其他同產業在儲存設備系統採購及規劃上參考的依據，並可以讓整體系統在上線後，不會有因為儲存設備系統效能差而影響產線或研發的問題發生。 ;In the semiconductor industry, electronic design automation tools are used for designing the integrated circuits. The infrastructure of design automation tools includes computing nodes, network and storage system. As the IC design improves continuously, the data volume on a wafer increases with the same ratio. In order to accelerate the IC design and shorten the production time, the integration of electronic design automation software with its infrastructure becomes important. The purpose of this study is to find out the suitable parameters and configuration for the computing environment of electronic design automation so as to enhance the overall performance the overall system configuration. The case - R Company is an IC design company. It focuses on IC research and design. Its product lines span communications network ICs, computer peripheral ICs and multimedia ICs. In-depth interviews were applied to understand R Company’s EDA tool software environment and storage system architecture. R Company′s real project data were used to test different storage system configurations in order to find out the best configuration for the company’s computing environment. The results of the tests reveal that the Flash Pool cache always performs better than the Flash Cache for the NetApp storage system. Also, when a NetApp storage system configured two data "aggregates" and more than four "volumes" under one controller, the project run time reduced significantly. Therefore, using Flash Pool cache with the above storage configuration can improve the overall performance of electronic design automation project. The results of this study can be a guideline for other industries in procuring and planning storage systems. Applying the results will improve the overall system performance by reducing the performance problem caused by the storage system, avoiding the delay of the IC design and production.
