Reconfigurable asynchronous pipelines: From formal models to silicon

Sokolov, Danil; Gennaro, Alessandro de; Mokhov, Andrey

doi:10.23919/date.2018.8342264

Cited by 4 publications

(4 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Parser combinators by Swierstra and Duponcheel [1996] paved the way to the Alternative type class (ğ7.2). Yallop [2010] proposed to extend Applicative with a method of type f Bool -> f a -> f a -> f a łfor capturing computations where control flow is dynamic, but dataflow is staticž; similar ideas were studied by hardware designers in the context of synchronous [Dennis and Misunas 1975] and asynchronous [Mokhov 2009;Sokolov et al 2018] control circuits. We have also found early online discussions [Permyakov et al 2012;Yorgey et al 2009] that searched for type classes like Selective but did not progress further.…”

Section: Related Workmentioning

confidence: 99%

Selective applicative functors

Mokhov

Lukyanov

Marlow

et al. 2019

Proc. ACM Program. Lang.

Self Cite

View full text Add to dashboard Cite

Applicative functors and monads have conquered the world of functional programming by providing general and powerful ways of describing effectful computations using pure functions. Applicative functors provide a way to compose independent effects that cannot depend on values produced by earlier computations, and all of which are declared statically. Monads extend the applicative interface by making it possible to compose dependent effects, where the value computed by one effect determines all subsequent effects, dynamically.This paper introduces an intermediate abstraction called selective applicative functors that requires all effects to be declared statically, but provides a way to select which of the effects to execute dynamically. We demonstrate applications of the new abstraction on several examples, including two industrial case studies.Consider a simple example, where we use the monad f = IO to describe an effectful program that prints "pong" to the terminal if the user enters "ping": pingPongM :: IO () pingPongM = getLine >>= \s -> if s == "ping" then putStrLn "pong" else pure ()The first argument of the bind operator reads a string using getLine :: IO String, and the second argument is the function of type String -> IO (), which prints "pong" when s == "ping".As we will see in sections ğ3 and ğ4, in some applications it is desirable to know all possible effects statically, i.e. before the execution. Alas, this is not possible with monadic effect composition. To inspect the function \s -> ..., we need a string s, which becomes available only during execution. We are therefore unable to predict the effects that pingPongM might perform: instead of conditionally executing putStrLn, as intended, it might delete a file from disk, or launch proverbial missiles.Applicative functors, introduced by McBride and Paterson [2008], can be used for composing statically known collections of effectful computations, as long as these computations are independent from each other. The key ingredient of applicative functors is the apply operator, denoted by <*>:The operator takes two effectful computations, which Ð independently Ð compute values of types a -> b and a, and returns their composition that performs both computations, and then applies the obtained function to the obtained value producing the result of type b. Crucially, both arguments and associated effects are known statically, which, for example, allows us to pre-allocate all necessary computation resources upfront (ğ3) and execute all computations in parallel (ğ4).Our ping-pong example cannot be expressed using applicative functors. Since the two computations must be independent, the best we can do is to print "pong" unconditionally:pingPongA :: IO () pingPongA = fmap (\s -> id) getLine <*> putStrLn "pong"Here we use fmap (\s -> id) to replace the input string s, which we now have no need for, with the identity function id :: () -> (), thus matching the type of putStrLn "pong" :: IO (). We cannot execute the putStrLn "pong" effect conditionally but, on the posit...

show abstract

Section: Related Workmentioning

confidence: 99%

Selective applicative functors

Mokhov

Lukyanov

Marlow

et al. 2019

Proc. ACM Program. Lang.

Self Cite

View full text Add to dashboard Cite

show abstract

“…The first set of benchmarks includes an on-chip power management controller of a buck converter [33] and an asynchronous controller for the reconfigurable pipeline of a dataflow processor [34].…”

Section: Ad-hoc Controllersmentioning

confidence: 99%

“…Its controller manages the energy quality of the result by controlling the number of active pipeline stages. It is an important case study, as it was fabricated in an application specific integrated circuit and tested [34]. Three of its 13 scenarios that compose the reconfigurable pipeline are specified below in the text‐form, i.e.…”

Section: Algorithm and Tool Validationmentioning

confidence: 99%

See 1 more Smart Citation

Efficient composition of scenario‐based hardware specifications

Gennaro

Stankaitis

Mokhov

2019

IET Computers & Digital Techniques

Self Cite

View full text Add to dashboard Cite

Complex hardware systems can be designed by breaking down their behaviour into high-level descriptions of constituent scenarios and then composing these scenarios into an efficient hardware implementation using a form of highlevel synthesis. There are a few existing methodologies for such scenario-based specification and synthesis, and in this paper we focus on highly concurrent systems, whose scenarios are typically described using explicit concurrency models such as partial orders. We propose a new algorithm for composition of partial order scenarios. Unlike previously published methods, the proposed algorithm supports composition constraints, which allow the designer to restrict certain aspects of the composition in order to reuse legacy IP. Furthermore, our implementation is more scalable and can cope with specifications comprising hundreds of scenarios at the cost of only 5% of area overhead compared to optimal solutions obtained by exhaustive search. The proposed algorithm is implemented in an open-source EDA tool, validated on a set of benchmarks, and compared to the state-of-the-art behavioural composition approaches and to other existing methodologies that make use of behavioural synthesis.

show abstract