A stream compiler for communication-exposed architectures

Abstract

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 softwareexposed communication between one unit and another (e.g., Raw, SmartMemories, TRIPS). However, for their use to be widespread, it will be necessary to develop compiler technology that enables a portable, high-level language to execute efficiently across a range of wire-exposed architectures.In this paper, we describe our compiler for StreamIt: a high-level, architecture-independent language for streaming applications. We focus on our backend for the Raw processor. Though StreamIt exposes the parallelism and communication patterns of stream programs, some 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.We have implemented a fully functional compiler that parallelizes StreamIt applications for Raw, including several load-balancing transformations. Using the cycle-accurate Raw simulator, we demonstrate that the StreamIt compiler can automatically map a high-level stream abstraction to Raw without losing performance. We consider this work to be a first step towards a portable programming model for communication-exposed architectures.

Scalable framework for mapping streaming applications onto multi-GPU systems

“…The StreamIt programming language [8] is based on synchronous data flow graphs [18]. It has been compiled on a myriad of architectures including multi-cores [9], Cell [17], Raw [8] and even FPGAs [10].…”

“…It has been compiled on a myriad of architectures including multi-cores [9], Cell [17], Raw [8] and even FPGAs [10]. Moreover, there is previous research in coarse-grained mapping of StreamIt to GPU platforms [23].…”

“…Hence, several stream programming languages were proposed to ease the expression of parallelism in such applications [4,8]. These languages capture computation in the form of a stream graph, where the graph nodes can consist of selfcontained tasks, or finer-grained compute elements.…”

“…These languages capture computation in the form of a stream graph, where the graph nodes can consist of selfcontained tasks, or finer-grained compute elements. StreamIt [8] is a platform-independent stream programming language that distinguishes among the alternative options. It is suitable for compilation onto a large number of platforms due to the fine-grained parallelism it exposes.…”

Lossless Image Compression Exploiting Streaming Model for Efficient Execution on Multicores

“…From the viewpoint of predictive lossless image compression, increased computational capability of multicores can be efficiently used by parallel execution of parts of algorithms that are parallelizable and computationally demanding, notably adaptive predictor modeling. Moreover, streaming nature of these computations enable their implementation in a high-level streaming language such as StreamIt which exposes task, data and pipeline parallelism and enables portability and scalability, features not possible with old-fashioned programming practices which optimize in low level architecture-specific languages [13,14].…”