Vijay Karamcheti and Andrew A. Chien. View caching: Efficient software shared memory for dynamic computations. In Proceedings of the International Parallel Processing Symposium, pages 483--489, 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... It performs a number of sequential optimizations such as access region expansion, object inlining [10] and method inlining to optimize object oriented code [20] In addition, the compiler supports concurrent locality based optimizations, such as dynamic pointer alignment [26] and view caching [14]. In both the sequential and parallel realms, these features make the Concert compiler state of the art for object oriented compilers. 1.5 Summary of Results Using our system on the SGI Cray Origin 2000 shared memory machines, we have achieved good speedups for our SAMR application. Sequential ....

Vijay Karamcheti and Andrew A. Chien. View caching: Efficient software shared memory for dynamic computations. In Proceedings of the International Parallel Processing Symposium, 1997.

....such as data transfer time. More important than emulation, a low level model offers the promise that network traffic and protocol overhead may be reduced over that required by a higher level model by programmer specialization [78] or through automatic, compile time analysis and specialization [23, 34]. An ideal low level model provides a complete set of communication operations and exposes fundamental costs. The programmer is thus given the ability to craft communication protocols tailored to the application and to minimize communication costs using application specific knowledge. The ....

Vijay Karamcheti and Andrew A. Chien. View Caching: Efficient Software Shared Memory for Dynamic Computations. In Proceedings of the 11th International Symposium on Parallel Processing, April 1997.

....expresses restrictions on object usage. We also describe the design and implementation of a translator, which converts VJava programs into base Java that uses JNI calls to access a software shared memory layer that provides composable primitives for building custom coherence protocols [11]. The run time environment involves independent JVMs running on each of the nodes of a distributed cluster of workstations, sharing objects using the software shared memory layer. To assess the convenience of using these extensions and quantify their performance impact, we have also built a ....

....a thin JNI (Java Native Interface) wrapper on top of the C based software DSM layer to enable VJava objects to interface with the corresponding shared regions. The DSM library that we used, called VCache, provides an efficient implementation of entry consistency protocols for shared regions [11,16]. To supply a bootstrap mechanism for shared applications, we implemented a simple name server that registers names of global objects at a known address in the Java RMI Registry; joining nodes can look up the remote reference using this name server. Figure 5 shows the overall architecture. When ....

Vijay Karamcheti and Andrew A. Chien. View caching: Efficient software shared memory for dynamic computations. In Proceedings of the International Parallel Processing Symposium, pages 483--489, 1997.

....to create regions of optimized sequential code with the efficiency of a sequential uniprocessor implementation. 4. 3 Locality Optimizations Since global pointer based data structures are fundamental for many dynamic (e.g. data dependent) computations, Concert supports two locality optimizations [21, 37] to efficiently implement such structures on modern architectures with deep memory hierarchies, such as NUMA machines, whether cache coherent or not. When static coarse grained aliasing information is available, we apply dynamic pointer alignment, a generalization of static loop tiling and ....

....actually light weight threads at compile time from loop bodies and function calls. At run time, the program concurrency structure allows these iterations to be reordered dynamically, guided by runtime data access information, to maximize data reuse and hide communication latency. View caching [21] supports efficient runtime object caching in dynamic computations, relying on application knowledge of data access semantics to construct customized latency tolerant coherence protocols that require reduced message traffic and synchronization. Application knowledge is used to infer information ....

Vijay Karamcheti and Andrew A. Chien. View caching: Efficient software shared memory for dynamic computations. In Proceedings of the International Parallel Processing Symposium, 1997.

.... [17, 18] ffl Directed cloning and optimization (procedure and object inlining) 6, 20] ffl Compiler managed locality and memory latency management [27] ffl Efficient, robust communication primitives [12, 14] ffl Hybrid stack heap execution (efficient multithreading) 14, 19] ffl View Caching [13] 1.2 Application Suite In this paper, we use a suite of seven irregular applications to evaluate parallel programming support in ICC . Table 1 briefly describes the applications. Although spanning diverse computational domains, these applications share common characteristics that make them ....

Vijay Karamcheti and Andrew A. Chien. View caching: Efficient software shared memory for dynamic computations. In Proceedings of the International Parallel Processing Symposium, 1997.

....to create regions of optimized sequential code with the efficiency of a sequential uniprocessor implementation. 3. 3 Locality Optimizations Since global pointer based data structures are fundamental for many dynamic (e.g. data dependent) computations, Concert supports two locality optimizations [25, 43] to efficiently implement such structures on modern architectures with deep memory hierarchies, such as NUMA machines, whether cache coherent or not. When static coarse grained aliasing information is available, we apply dynamic pointer alignment, a generalization of static loop tiling and ....

....actually light weight threads at compile time from loop bodies and function calls. At run time, the program concurrency structure allows these iterations to be reordered dynamically, guided by runtime data access information, to maximize data reuse and hide communication latency. View caching [25] supports efficient runtime object caching in dynamic computations, relying on application knowledge of data access semantics to construct customized latency tolerant coherence protocols that require reduced message traffic and synchronization. Application knowledge is used to infer information ....

V. Karamcheti and A. A. Chien. View caching: Efficient software shared memory for dynamic computations. In Proceedings of the International Parallel Processing Symposium, 1997.

....range of irregular applications. Our application study is done in the context of the Illinois Concert system, a high performance compiler and runtime for parallel computers which has been the vehicle for extensive research on compiler optimization and runtime techniques over the past five years [9, 32, 29, 30, 33, 28, 31, 12, 44, 23, 22]. While no system contains all known optimizations, the Concert system contains a wide range of aggressive optimizations, and has been used to demonstrate high performance in absolute terms on a wide range of applications [23, 45, 10] In effect, the Concert system automatically addresses many of ....

....) 2. 3 Illinois Concert System Our application study is done in the context of the Illinois Concert system, a high performance compiler and runtime for parallel computers which has been the vehicle for extensive research on compiler optimization and runtime techniques over the past five years [9, 32, 29, 30, 33, 28, 31, 12, 44, 23, 22]. While no system contains all known optimizations, the Concert system contains a wide range of them, and has been used to demonstrate high performance in absolute terms on a wide range of applications [23, 45, 10] In effect, the Concert system automatically addresses many of the concerns which ....

[Article contains additional citation context not shown here]

Vijay Karamcheti and Andrew A. Chien. View caching: Efficient software shared memory for dynamic computations. In Proceedings of the International Parallel Processing Symposium, 1997.

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