Synchronization constraints

HIGH LEVEL SYNTHESIS OF ASICs UNDER TIMING AND SYNCHRONIZATION CONSTRAINTS, D. C. Ku, G. De Micheli ISBN 0-7923-9244-2... 

In order to reduce non-determinism and to increase the effectiveness of testing, non-preemptive execution is sometimes desirable (Bums 2001). The standard way of implementing many high-integrity applications is with a cyclic executive (Baker Shaw 1989). Using this technique a sequence of procedures is called within a defined time interval. Each procedure mns to completion and there is no concept of preemption. Data is passed from one procedure to another via shared variables and no synchronization constraints are needed, since the procedures never run concurrently. The major disadvantage with non-preemption is that it will usually (although not always) lead to reduced schedulability. 

This modeling relies on the existence of a runtime system whose behavior is known. The response time of the system is guaranteed by the functioiung of the host stracture. At the onset of an external event, the reaction time of the system (execution cycle of the mntime) and the spread to all modules in all levels of the system are determined. The introduction of an additional synchronization constraint between processes or the taking into acconnt of the process execntion time or commuiucation time between modules is urmecessary. The operation of the mntime ensures that the output calculations are carried out before the arrival of new input. The operation is cyclic and deterministic. 

Synchronization constraints. Concurrently executing modules need to synchronize with one another to coordinate their actions. Synchronization is especially important if the design is to communicate with its environment... 

Perform scheduling uruier detailed timing and synchronization constraints. The resulting schedule is used to synthesize the control unit. 

Relative control optimization a technique, called control resynchronization, that reduces the control implementation costs while still satisfying the required timing and synchronization constraints. 

The main algorithmic contributions of this research are described in the next four chapters. Ch ter 6 presents the relative scheduling formulation that includes description of the algorithms and analysis of their prqterties. Chapto 7 describes conflict resolution under timing constraints. Chapter 8 describes the generation of the control circuit from a relative schedule. Chapter 9 describes the control resynchronization optimization that reduces the area of the control implementation under timing and synchronization constraints. 

As stated earlier, the primary goal of this research is to develq) methods of synthesizing hardware from abstract specifications under both detailed timing and synchronization constraints. Detailed timing constraints ctg)ture minimum and maximum bounds on the start time of operations synchronization constraints model handshaking and coordination among concurrent computation threads, and are represented as operations with data-dependent execution delays. 

HIGH LEVEL SYNTHESIS OF ASICs UNDER TIMING AND SYNCHRONIZATION CONSTRAINTS... 

High level synthesis of ASICs under timing and synchronization constraints / by David C. Ku and Giovanni De Micheli. 

For concreteness, let us suppose that the universe has a temporal depth of two to accommodate a Fi edkin-type reversibility i.e. the present and immediate past are used to determine the future, and from which the past can be recovered uniquely. The RUGA itself is deterministic, is applied synchronously at each site in the lattice, and is characterized by three basic dimensional units (1) digit transition, D, which represents the minimal informational change at a given site (2) the length, L, which is the shortest distance between neighboring sites and (3) an integer time, T, which, while locally similar to the time in physics, is not Lorentz invariant and is not to be confused with a macroscopic (or observed) time t. While there are no a priori constraints on any of these units - for example, they may be real or complex - because of the basic assumption of finite nature, they must all have finite representations. All other units of physics in DM are derived from D, L and T. 

Figure 7.3), leading to the following adapted constraints for the synchronization of the storages ... 

An important constraint in carrying out virtual displacements is that the individual mole numbers n must change in proper synchronization to accord with the chemical reaction 7 i/jA - 0. This can be guaranteed by introducing an Infinitesimal, virtual (which also implies "reversible") unit of advancement, SA, of the chemical reaction S i Ai - 0, such that the change in mole number of every species i is given by 6 - i SA. Then Eq. (2.9.1) reads... 

---