Database Performance Evaluation Methodology for Design Automation Systems: A Case Study of MICON

Design Automation Systems such as MICON rely heavily on database systems for design information retrieval. In our current configuration, over 90% of the system synthesis time is spent on accessing the database. Here we address the significance of the database system by considering three database issues. First, a way to compare the performance of different database systems while embedded in a computer aided design (CAD) system. Second, a way to predict database performance when synthesizing different designs. Finally, a way to identify key areas for improvement resulting in improved performance of the CAD system. The methodology was applied to two different relational databases coupled to the MICON system. The results was over a factor of three speedup in total database processing time resulting in over a factor of two reduction in MICON run time. 1.0 Introduction Database systems are found in a variety of computer applications. Not surprisingly, Design Automation (DA) tools are among those applications which require the ability to store and rapidly access large amounts of data. One such design automation system, MICON [Birmingham, Gupta, Siewiorek 1992], synthesizes designs at the system level. The set of MICON tools allow for the capture of hardware design knowledge and the synthesis of new systems based on its accumulated knowledge. The synthesis tool, M1, is a knowledge-based system which designs by the method of synthesis-by-composition. Figure 1 shows a sample M1 input and output. Synthesis-by-composition involves traversing a hierarchy of objects and putting together objects that fit the required criteria. The selection of objects relies heavily on retrieval of information from the database. Section 1 of this report describes the relationship of the database to the MICON synthesis system while Section 2 presents measurements on time spent in various phases of the design of a personal computer and a process control computer. Section 3 presents a performance model for databases inspired by instruction frequency performance models commonly used to describe computer workstations. Section 4 demonstrates the use of the performance model in predicting database performance. Section 5 illustrates how simple modifications can be made to remove performance bottlenecks yielding over a factor of two improvement in system performance. Finally, section 6 draws some conclusions about the research. 1.1 Database Organization The database which MICON uses is built upon the table-based relational model [Korth, Silberschatz 1991]. Each row in a table represents a relationship among the values in the row. Expressions which identify sets of information are formulated on relational algebra and relational calculus. Expressions for sets of information are converted into database queries (requests for information). SQL (Structured Query Language) is an ANSI standard that is based on both relational algebra and relational calculus. Database Performance Evaluation Methodology for Design Automation Systems: A Case Study of MICON 1 Query Tot.~ amount of boa~ area ~ ~tuarc inche~ > Total ammlnt of board co$~ < nearest dollar > To~l ~lnount of power di~sipatlon aIIowcd < milliwatts ~ Is a CO.PROCESSOR needed Is HO needed Is SIO needed Is TIMER needed Is a Keyboard Commoner needed Namo f Pwcessor Fam~y Pmcessor na n~ i000