Target architecture

This review concerns the synthesis and characterization of octa-arm polyisobutylene (PIB) stars, allyl-terminated octa-arm PIB stars, and octa-arm star blocks by using a novel octafunctional caHx[8]arene-based initiator 1. Scheme 1 shows the structure of 1 and the target architectures. The syntheses were carried out under living carbocationic polymerization conditions. 

The target architecture aimed at is illustrated in Fig. 7.27. Each layer depends on the one below and realizes the prerequisites for the layer above ... 

In this case, the self-assembly process precedes the final modifications that lead to the targeted architecture. Using... 

Domain-specific synthesis tools which fully exploit the characteristics of the target architecture style and the corresponding application domain. The tools will frequently differ for the alternative target styles. For instance, a scheduler that is efficient for one style may be unsuitable or inefficient for another. 

By definition, the dimension of the DG of an algorithm is identical to the index space dimension. In particular, the bit-level algorithm of equation (14) should be described by a W-D DG. However, the construction of the multidimensional DG is avoided by using combinatorial logic to compute the values of Zk -)j thus reducing the problem to multiple-operand binary addition, which can be represented by a 2-D DG. Moreover, in this DG, the properties of the target architecture may be embodied. 

This target application domain features a number of important characteristics which are heavily exploited in our target architecture style and our synthesis approach [1, 18, 8] ... 

All the synthesis steps mentioned above can be executed either automatically or interactively. The programmed ASIC emulator can be inserted directly into the overall mechatronic system to facilitate real-time test runs. In the following, all main synthesis steps will be described in the same top-down sequence as they are implemented through the synthesis script. Throughout the rest of the chapter, the terms ASIC emulator (board), rapid prototyping board, and target architecture will be used synonymously. 

Component placement places the registers and FUs in order to reduce the connections. Because Amical uses a bus-based target architecture, the placement of registers and functional units is very important. Optimal placement will minimize the number of busses necessary. 

Optimize the result of step 2 to produce a result well suited to the implementation architecture. The goal of the optimization and the strategies involved depend strongly on the target architecture. For instance, EDA took for FPGAs with limited interconnections seek to reduce the number of block interconnects. 

Optimality. A high degree of optimality can be achieved through a careful definition of the target architectural style and application domain to be supported by the compiler, and the subsequent exploitation of both architectural and algorithmic properties in the synthesis methodology and tools. 

In Section 1, it has been motivated that a relation exists between the target architectural style and the corresponding target application domain. For each particular style, both the architectural and the application features should be exploited in order to arrive at an efficient interactive synthesis methodology which can compete with manually derived designs. As a consequence, this has an effect on both the sequence of synthesis tasks to be performed, and on the choice of synthesis techniques used to solve the individual (sub)tasks. These two factors combined are called the synthesis script henceforth. Because of the strong relation with the architectural styles, we have opted for very customised scripts. 

In this paper, the novel CATHEDRAL environment has been presented which supports important high-level synthesis tasks for mapping complex, irregular signal processing applications into efficient architectures. Two scripts have been presented, each tuned towards a different application domain with a corresponding target architectural style. They are differentiated based on the ratio between... 

In this chapter, we will present the processor synthesis part of THEDA. The general design consideration is described in Section 2. The target architecture is described in Section 3. Scheduling is described in Section 4. Constructive and iterative refined approaches to the data path binding problem are described in Section 5. Finally, we conclude with a summary. 

The bus assignment phase is only necessary if the target architecture for the data path is bus oriented. In this case the previous unidirectional connections either from register to functional unit or vice versa are merged into bidirectional busses. For this step each data transfer must be treated separately and be mtqtped onto a bus. 

---