<p>Streaming applications like multimedia and radar signal processing applications are becoming </p><p>increasingly compute-intensive. To overcome the computational demands new parallel </p><p>architectures are emerging. </p><p> </p><p>The programming tools provided with these architectures require low-level programming, which </p><p>creates a need for a common high-level architecture independent language that can exploit </p><p>parallelism efficiently. One such language is StreamIt, designed around the notions of streams </p><p>and stream transformers, which allows efficient mapping to a variety of architectures. </p><p> </p><p>The overall goal of this master’s thesis is to evaluate the StreamIt language from a </p><p>programmability and portability point of view. An MPD-application has been developed in </p><p>StreamIt, which is executed on the RAW simulator. Furthermore, a code generator is designed to </p><p>compile and execute the application on the XPP simulator. </p><p> </p><p>The conclusions drawn are that StreamIt is easy to learn, but hard to use because of its </p><p>programming paradigm as compared to conventional languages. StreamIt programming involves </p><p>thinking in terms of streams instead of globally accessed memory. The structure of StreamIt </p><p>makes it easy to reuse components and modify the application. The construction of the compiler </p><p>makes it possible to port StreamIt application to various architectures.</p>

Streaming applications like multimedia and radar signal processing applications are becoming increasingly compute-intensive. To overcome the computational demands new parallel architectures are emerging.   The programming tools provided with these architectures require low-level programming, which creates a need for a common high-level architecture independent language that can exploit parallelism efficiently. One such language is StreamIt, designed around the notions of streams and stream transformers, which allows efficient mapping to a variety of architectures.   The overall goal of this master’s thesis is to evaluate the StreamIt language from a programmability and portability point of view. An MPD-application has been developed in StreamIt, which is executed on the RAW simulator. Furthermore, a code generator is designed to compile and execute the application on the XPP simulator.   The conclusions drawn are that StreamIt is easy to learn, but hard to use because of its programming paradigm as compared to conventional languages. StreamIt programming involves thinking in terms of streams instead of globally accessed memory. The structure of StreamIt makes it easy to reuse components and modify the application. The construction of the compiler makes it possible to port StreamIt application to various architectures.

StreamIt is a language specifically designed for modern streaming applications. A certain important class of these applications operates on streams of bits. This paper presents the motivation for a bit-packing optimization to be implemented in the StreamIt compiler for the RAW Architecture. This technique aims to pack bits into integers so that operations can be performed on multiple bits at once thus increasing the performance of these applications considerably. This paper gives some simple example applications to illustrate the various conditions where this technique can be applied and also analyses some of its limitations. / Singapore-MIT Alliance (SMA)

No description available.

With the increasing miniaturization of transistors, wire delays are becoming a dominant factor in microprocessor performance. To address this issue, a number of emerging architectures contain replicated processing units with software-exposed communication between one unit and another (e.g., Raw, SmartMemories, TRIPS). However, for their use to be widespread, it will be necesary to develop a common machine language to allow programmers to express an algorithm in a way that can be efficiently mapped across these architectures. We propose a new common machine language for grid-based software-exposed architectures: StreamIt. StreamIt is a high-level programming language with explicit support for streaming computation. Unlike sequential programs with obscured dependence information and complex communication patterns, a stream program is naturally written as a set of concurrent filters with regular steady-state communication. The language imposes a hierarchical structure on the stream graph that enables novel representations and optimizations within the StreamIt compiler. We have implemented a fully functional compiler that parallelizes StreamIt applications for Raw, including several load-balancing transformations. Though StreamIt exposes the parallelism and communication patterns of stream programs, analysis is needed to adapt a stream program to a software-exposed processor. We describe a partitioning algorithm that employs fission and fusion transformations to adjust the granularity of a stream graph, a layout algorithm that maps a stream graph to a given network topology, and a scheduling strategy that generates a fine-grained static communication pattern for each computational element. Using the cycle-accurate Raw simulator, we demonstrate that the StreamIt compiler can automatically map a high-level stream abstraction to Raw. We consider this work to be a first step towards a portable programming model for communication-exposed architectures. / Singapore-MIT Alliance (SMA)