J. McGraw, S. Skedzielewski, R. Oldehoeft, J. Glauert, C. Kirkham, B. Noyce, R. Thomas. "SISAL: Streams and Iteration in a Single Assignment Language." Language Reference Manual, Version 1.2, M146 Rev. 1, University of CaliforniaDavis, March 1985.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....Theorem 6.1.1 Given an SDSP MCP PN G which models I clean pipelines and contains n SDSP transitions, the computation rate of any SDSP transition in G can never be greater than I=n, i.e. fl I=n. 6. 2 Simulation Results A set of Livermore Loops was chosen for the study; all were written in SISAL [12, 24]. Our simulations were performed on a compiler simulator testbed developed at McGill University [16] The testbed consists of a prototype SISAL compiler capable of producing dataflow code (known as A Code) 30, 31] For this particular study, we modified the simulator to permit analysis of cyclic ....

J. R. McGraw and et al. SISAL: Streams and iteration in a single assignment language--- language reference manual version 1.2. Technical Report M-146, Lawrence Livermore National Laboratory, 1985.

....approach more flexibility is gained, many of the flaws mentioned in the previous approach remain. A different approach is to use a programming paradigm that does not obscure parallelism and leaves parallelization to the compiler. Examples are dataflow languages (e.g. ID Nouveau [Arvind88] SISAL [Sked85]) and functional languages (Haskell [Hudak89] Miranda [Turner 85] Crystal [Chen86] A problem regarding the execution efficiency of those languages is a.o. the single assignment nature, yielding much overhead when manipulating large data structures. Compilers for functional programs use graph ....

S. Skedzielewski, J. McGraw, S. Allen, R. Oldehoeft, J. Glauert, "SISAL : Streams and Iteration in a Single Assignment Language," Report Lawrence Livermore National Laboratory , M - 146, March 1985.

....data types for ecient use of FPGA space; and user control of optimizations. SA C draws ideas from a number of languages. The general syntax is derived, as much as possible, from C; powerful multi dimensional array capability is drawn from Fortran 90; SA C s loop generators are inspired by Sisal [9]. Data types in SA C include signed and unsigned integers and xed point numbers, with user speci ed bit widths. For example, a uint8 is an 8 bit unsigned type, whereas a x12.4 is a xed point type with a sign bit, seven whole number bits and four fractional bits. SAC also has oat and double ....

J. McGraw and et al. SISAL: Streams and Iteration in a Single Assignment Language: Reference Manual Version 1.2. Lawrence Livermore National Laboratory, memo m-146 rev. 1 edition, 1985.

....values as outputs. When executing a functional language program, all functions whose inputs are available can be evaluated in parallel without fear of interference. This data driven parallel 18 execution model is the foundation of a number of functional languages like VAL [2] ID [34] and SISAL [15]. An alternative execution model called lazy evaluation evaluates a function only if its inputs are available and it has been determined that the result of the function is required to produce the output of the program. Lazy evaluation permits the programmer to define and use infinite data objects ....

J. McGraw et al. Sisal: Streams and iterations in a single-assignment language. Technical Report M-146, Lawrence Livermore Noational Laboratories, March 1985.

....cache consistency problem arises if the memory model supports synchronizing memory operations. A multiprocessor that implements such a program execution model can support all determinate computation, including, for example, all scientific computations expressible in the Sisal programming language [McGraw 85, Cann 92] As we have shown elsewhere [Dennis Gao 95] a general class of nondeterminate computations can be supported by adding a single special memory operation on non cached locations, without introducing any additional consistency requirement to the memory model. These circumstances raise ....

J. McGraw, S. Skedzielewski, S. Allan, R. Oldehoeft, J. Glauert, C. Kirkham, B. Noyce, and R. Thomas. SISAL: Streams and Iteration in a Single Assignment Language: Reference Manual Version 1.2. Technical Report M-146, Rev. 1, Lawrence Livermore National Laboratory, 1985.

....Code from Functional Programs 3 1 Introduction Because of the elegance and power of functional languages, one would expect them to be naturally suited for numeric and scientific computing. However, this potential has not yet been well exploited, in spite of efforts such as SISAL and others [10, 9]. To fully exploit this potential, compiled functional programs must run as efficiently as compiled imperative programs. Two of the difficulties encountered in compiling functional languages seem to be avoiding the high overhead of context switching and dealing with arrays efficiently. We have ....

.... 1 =i =N of real; double r[3] #define r(i) r[ i) 1] y : i 1 =i =N of real; double y[3] #define y(i) y[ i) 1] Z : t1,t2 (t2,2) t1 =N; 1 =t2 of real; double Z[6] #define Z(t1,t2) Z[3 (t1) t2) 7] Y : t1,t2 t2 N =t1 =N; N 1 =t2 of real; double Y[9]; #define Y(t1,t2) Y[3 (t1) t2) 7] B : t1,t2 1 =t1 =N; t2=0 of real; double B[3] #define B(t1,t2) B[ t1) 1] main( int i, t1, t2; for (i=1; i =N; i ) printf( Input r[ d] i) gets(s) r(i) atof(s) B(1,0) 1.0; Y(1,4) r(1) for(t1=2; t1 =N; t1 ) ....

J.R. McGraw, S.K. Skedzielewski, S. Allan, and D. Grit. SISAL---streams and iteration in a single--assignment language. Language Reference Manual, Version 1.2, Jan 1985.

....size. When compared to the original factor of 5.6 without padding enabled, program runtimes are much more predictable. 4 Related Work In high level functional programming languages, lists rather than arrays form the predominant data structure. The most prominent exception is the language Sisal [10]. However, Sisal represents arrays as vectors of vectors rather than as contiguous data which renders array padding useless. So, we are not aware of any similar optimization technique in this area. However, in high performance computing, mostly based on Fortran, data locality has long been identi ....

J.R. McGraw, S.K. Skedzielewski, S.J. Allan, R.R. Oldehoeft, et al. Sisal: Streams and Iteration in a Single Assignment Language: Reference Manual Version 1.2. M 146, Lawrence Livermore National Laboratory, LLNL, Livermore California, 1985.

....dependencies of the program. This also means that control can be distributed. Further, no central memory system is inherent to the model since data is not written in by any instruction but is passed from one function to the next. Sisal (Stream and Iteration in a Single Assignment Language) [20] is such a functional language which was originally designed by collaborating teams from the Lawrence Livermore National Laboratory, Colorado State University, 1 the University of Manchester and Digital Equipment Corporation. The goal of the project was to design a generalpurpose implicitly ....

J. McGraw, S. Skedzielewski, S. Allan, D. Grit, R. Oldehoeft, J. Glauert, I. Dobes, and P. Hohensee. SISAL-Streams and Iterations in a Single Assignment Language, Language Reference Manual, version 1. 2. Technical Report TR M146, University of California - Lawrence Livermore Laboratory, March 1985.

....in the figure, they are used for the digits to flow as the sequences of intervals are generated whereas the backward links are used to control the precision through backtracking. The program internal image is close to the one which would have been built from a single assignment language [7]. As an instance of an operator is created in memory, all the digits of the inputs deduced from the initial accuracy provided by the user are available. The operator treats as many digits as possible according to the considerations developed in the next sub section. The digits of the result are ....

James McGraw et al. SISAL: streams and iteration in a single assignment language - version 1.2. Language reference manual, Lawrence Livermore National Laboratory, 1985. 10

.... of related problems such as task partitioning and scheduling [1, 7, 20, 3, 6] task sequencing [29] and instruction reordering [21] In the area of functional languages, solutions were proposed both for non strict (lenient) languages such as Id [16, 17] and for strict languages such as Sisal [15]. These partitioning algorithms are based on a program intermediate representations called structured dataflow graph [27] that consist of a collection of basic blocks and interfaces. Dependence set partitioning algorithms find a correct partitioning by merging nodes together with the same ....

J. R. McGraw et al. SISAL: Streams and iteration in a single assignment language--- language reference manual version 1.2. Technical Report M-146, Lawrence Livermore National Laboratory, 1985. 22

....arrays, including array sections similar to those in Fortran 90. For example, int10 A[ declares a variable length two dimensional array of ten bit integers, and int10 A[0, is its first row. It has powerful loop generators and return operators similar to those in the Sisal language[5]. It has multiple value returns and assignments. The elimination of pointers and recursion and the singleassignment restriction enable important compiler code optimizations. In compensation for these restrictions, there are powerful high level constructs to create and access arrays in concise ....

J. McGraw et.al., SISAL: Streams and Iteration in a Single Assignment Language: Reference Manual Version 1.2, Lawrence Livermore National Laboratory, Memo M-146, Rev. 1, 1985.

....of those places where the name is the input of an operation. Because operations execute only when all of their inputs are present, communication is always unsynchronized. Dataflow languages have taken different approaches to expressing repetitive operations. Lan23 guage such as Id [77] and Sisal [147, 148, 178] are first order functional (or single assignment) languages. They have syntactic structures looking like loops, which create a new context for each execution of the loop body (so that they seem like imperative languages except that each variable name may only be assigned to once in each ....

J. McGraw, S. Skedzielewski, S. Allan, R. Oldehoeft, J. Glauert, C. Kirkham, B. Noyce, and R. Thomas. Sisal: Streams and iteration in a single assignment language: Reference manual 1.2. Technical Report M-146, Rev.1, Lawrence Livermore National Laboratory, March 1985.

....for multithreaded execution models has been mainly done for functional languages within the data flow community. Two types of thread partitioning problems have been studied, one is targeted for non strict (lenient) languages such as Id [94, 97] and the other for strict languages such as Sisal [89]. Thread partitioning for non strict languages has more constraints than for strict languages, such as avoiding deadlock. Many partitioning algorithms have been proposed for such functional programming languages. All these partitioning algorithms are based on a program intermediate ....

J. R. McGraw et al. SISAL: Streams and iteration in a single assignment language---language reference manual version 1.2. Technical Report M-146, Lawrence Livermore National Laboratory, 1985.

....a dataflow graph, in favour of a multi threaded style where a sequence of dataflow instructions is treated as a sequential thread. Thus the compiler produces code not as a dataflow graph but as a collection of sequential threads. SISAL SISAL (Streams and Iteration in a Single Assignment language) [McGr84]] is a functional programming language designed to express algorithms, mainly in the area of numeric computation, for execution on highly parallel machines. Parallelism is detected automatically by the compiler which also decides on the amount of parallelism to exploit. However, SISAL was not ....

McGraw, J. R., et al. "SISAL: streams and iteration in a single-assignment language". Language Reference Manual, Ver.1.2, Lawrence Livermore National Laboratory M-146, 1984.

No context found.

J. McGraw, S. Skedzielewski, R. Oldehoeft, J. Glauert, C. Kirkham, B. Noyce, R. Thomas. "SISAL: Streams and Iteration in a Single Assignment Language." Language Reference Manual, Version 1.2, M146 Rev. 1, University of CaliforniaDavis, March 1985.

No context found.

J. McGraw, S. Skedzielewski, S. Allan, R. Oldehoeft, et al. Sisal: Streams and Iteration in a Single Assignment Language: Reference Manual Version 1.2. M 146, Lawrence Livermore National Laboratory, Livermore, California, USA, 1985.

No context found.

James McGraw, Stephen Skedzielewski, Stephen Allan, Rod Oldehoeft, John Glauert, Chris Kirkham, Bill Noyce, and Robert Thomas. SISAL: Streams and Iteration in a Single Assignment Language, Language Reference Manual Version 1.2. Lawrence Livermore National Laboratory, March 1985.

No context found.

McGraw, J., Skedzielewski, S., Allan, S., Oldehoeft, R., Glauert, J., Kirkham, C., Noyce, B., and Thomas, R. SISAL: Streams and Iteration in a Single Assignment Language, Language Reference Manual Version 1.2. Lawrence Livermore National Laboratory, Mar. 1985.

No context found.

J. McGraw et al. `Sisal: streams and iterations in a single assignment language, language reference manual, version 1.2', Technical Report M-146, LLNL, March 1985.

No context found.

J. McGraw et.al., SISAL: Streams and Iteration in a Single Assignment Language: Reference Manual Version 1.2, Lawrence Livermore National Laboratory, Memo M-146, Rev. 1, 1985.

No context found.

James McGraw, Stephen Skezielewski, Stephen Allan, Rod Oldehoeft, John Glauert, Chris Kirkham, and Robert Thomas. SISAL: Streams and Iterations in a Single Assignment Language. Lawrence Livermore National Laboratory, Reference manual version 1.2. manual M-146, Rev. 1 edition, March 1985.

No context found.

J. McGraw, S. Skedzielewski, S. Allan, R. Oldehoeft, J. Glauert, C. Kirkham, B. Noyce, and R. Thomas. SISAL: Streams and Iteration in a Single Assignment Language: reference manual version 1.2. Manual M-146, Rev. 1, Lawrence Livermore National Laboratory, Livermore, CA, March 1985.

No context found.

J. McGraw, S. Skedzielewski, S. Allan, D. Grit, R. Oldehoeft, J. Glauert, I. Dobes, and P. Hohensee. SISAL-Streams and Iterations in a Single Assignment Language, Language Reference Manual, version 1. 2. Technical Report TR M-146, University of California - Lawrence Livermore Laboratory, March 1985.

No context found.

James R. McGraw et al., "SISAL: Streams and Iteration in a Single Assignment Language---Language Reference Manual, version 1.2", Lawrence Livermore Nat. Lab. Tech. Rept., Mar. 1985 Rishiyur S. Nikhil and Arvind, "Can Dataflow Subsume von Neumann Computing?", Proc. 15th Ann.

No context found.

J. R. McGraw, S. K. Skedzielewski, S. J. Allan, R. R. Oldehoeft, J. Glauert, C. Kirkham, W. Noyce, and R. Thomas. SISAL: Streams and iteration in a single assignment language: Reference manual version 1.2. Manual M-146, Rev. 1, Lawrence Livermore National Laboratory, Livermore, CA, March 1985.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC