Asynchronous design methodologies: an overview
read more
Citations
A survey of research and practices of Network-on-chip
Razor: a low-power pipeline based on circuit-level timing speculation
Principles of Asynchronous Circuit Design: A Systems Perspective
Theory of latency-insensitive design
References
Petri nets: Properties, analysis and applications
Digital Systems Testing and Testable Design
Model for Delay Faults Based Upon Paths
Trace Theory for Automatic Hierarchical Verification of Speed-Independent Circuits
Related Papers (5)
Frequently Asked Questions (14)
Q2. How do the authors preserve the change diagram firing semantics?
To preserve the change diagram firing semantics, the I-Net transitions require one of the places to have at least three tokens (hence the three arcs from each place to one of the30transitions), and when fired the transition removes a net of one token from each place.
Q3. What is the main disadvantage of micropipelines?
While straight-line pipelines are easy to construct, pipelines with feedback or even more general dataflow can require complex control circuits, circuits micropipelines gives little help in synthesizing.
Q4. What is the main argument against delay-insensitive circuits?
In order to make delay-insensitive circuit design practical for general computations, the authors must create a set of basic modules that both obey delay-insensitive properties and give enough functionality to implement a wide class of circuits.
Q5. What is the argument on the limitations of delayinsensitive circuits?
Note that since elements are included which have more than one output (most notably the toggle element), the argument on the limitations of delayinsensitive circuits with only single-output gates does not apply.
Q6. What is the main power of module-based systems?
Although the authors have seen that module-based systems can ease manual design, their main power is seen when they are coupled with a high-level language and automatic translation software.
Q7. Why is the fundamental-mode assumption unreasonably restrictive?
This is because datapath elements tend to have multiple input signals changing in parallel, and the fundamental-mode assumption would unreasonably restrict the parallelism in datapath elements.
Q8. How many tokens can be used to mark a weak arc?
Since this change diagram is 2-bounded, the authors allow for up to two debts on each arc by marking each I-Net place corresponding to a weak precedence arc with two tokens initially.
Q9. How many gate delays are required to make a circuit stable?
Since this is only a separation of 2 gate delays between inputs (for transitions 4→5), this transition is 3 times faster than the required 6 gate delays.
Q10. What are the correctness constraints for a cyclic CD?
These correctness constraints include those previously mentioned, as well as connectedness (all transitions connected by some series of arcs) and switchover29correctness (transitions on a signal must alternate between “+” and “-” in all possible executions).
Q11. What is the limitation of Chu’s algorithm?
While Chu’s algorithm handles STG/ICs and some STG/NCs, its utility for automatic synthesis is limited by the large number of restrictions.
Q12. What is the way to avoid the chip boundary problem?
some approaches, including the communicating processes compilation method described later, restrict isochronic forks to small localized areas, avoiding the chip boundary problem.
Q13. What is the argument against delay-insensitive circuits?
As the above argument against most 2-input gates can easily be generalized to any n-input gate, where n≥2, delay-insensitive circuits with only single output gates can use only C-elements, single input gates (buffers and inverters), and wires.
Q14. What is the effect of separating circuit correctness?
This assumption also has the desirable effect of separating circuit correctness13from specific delays, so that delay optimization via transistor sizing and similar improvements can be applied without affecting circuit correctness.