SPEC. Standard performance evaluation corporation, 2003. www.specbench.org/benchmarks.html.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....code shape (such as George and Appel s environment of many copy instructions) etc. 2. 7 Our experiments In our experiments, we use a test suite of 169 routines, including code based on Forsythe et al. s book on numerical methods [23] the SPEC 90 benchmarks [40] and the SPEC 95 benchmarks [41]. The SPEC 95 benchmarks have been transformed by the insertion of advisory prefetch instructions intended to improve cache behavior [35] and this transformation has the impact of increasing the register requirements in those routines. Out of the suite of 169 routines, 59 required some amount of ....

SPEC release 1.10, September 1995. Standards Performance Evaluation Corporation.

....particular hardware configurations, code shape (such as George and Appel s environment of many copy instructions) etc. 2. 7 Our experiments In our experiments, we use a test suite of 169 routines, including code based on Forsythe et al. s book on numerical methods [23] the SPEC 90 benchmarks [40], and the SPEC 95 benchmarks [41] The SPEC 95 benchmarks have been transformed by the insertion of advisory prefetch instructions intended to improve cache behavior [35] and this transformation has the impact of increasing the register requirements in those routines. Out of the suite of 169 ....

SPEC release 1.2, September 1990. Standards Performance Evaluation Corporation.

....The individual assertions are quite small and fast (array bounds checks, testing for null, etc. but their performance impact can be significant. We wish to use random sampling to spread this cost among many users. We have applied sampling to CCured versions of several Olden [10] and SPECINT95 [23] benchmarks. All programs run to completion and we are simply measuring the overhead of performing the dynamic checks. 3.1.1 Whole Program Sampling Table 1 summarizes static aspects of the sampling transformation when applied to the entirety of each benchmark. For each program, we give the total ....

SPEC 95. Standard Performance Evaluation Corporation Benchmarks. <http://www.spec.org/osg/cpu95/CINT95/>, July 1995.

....the user s requirements, and purchase additional resources if not; alternatively task requirements, reservations and node resources may be specified in some generalised resource units. For example, the Eclipse [Bruno99] project measured processor bandwidth reservations in terms of SPECint95 [SPEC95] units; for a given platform, these units could be calibrated by running the standard SPECint95 benchmarks. The NodeOS [Calvert98] has proposed logarithmic standard RISC cycles to serve a similar purpose. 162 A further possibility would be to measure applications resources usage at run time ....

SPEC CPU95 Benchmark, 1995. Standard Performance Evaluation Corporation, http://www.spec.org/osg/cpu95/. (p 162)

....for this study. The simulator runs multiple Alpha executables through emulation. To put the maximum pressure on the memory subsystem, we are running a multiprogrammed workload so that no two threads share data or code. We chose eight of the most memory intensive applications from the SPEC95 suite [1] and SPLASH. Those applications are applu, apsi, arc2d, compress, gcc, hydro2d, ocean, flo52q. The applications are compiled with the DEC UNIX cc and f77 compilers using the highest levels of optimization. All of our studies assume a constant processor cycle time, adjusting access latencies and ....

Spec; standard performance evaluation corporation. In http://www.specbench.org, 1996.

....[60, 14, 28, 68] Despite modi cations, the Andrew benchmark has other limitations. It does not truly stress the I O subsystem since less than 25 of the time is spent performing I O [15] The Standard Performance Evaluation Corporation (SPEC) System File Server (SFS) 97 R1 V3.0 benchmark [74] is popular among the computer industry. It measures the performance of the le system running as an NFS server. This benchmark is not suitable for our use since it introduces complications and interference from NFS protocol processing and UDP IP network protocol trac. Bonnie, written by Tim Bray ....

SPEC. SFS 3.0. Whitepaper 1.0, Standard Performance Evaluation Corporation, Warrenton, Virginia, 2001.

....As well, human reading speeds may limit the maximum rate of user requests. A better model would have clients issue requests at a xed rate and simply increase the number of clients until throughput fails to increase any further or until response time is greater than some threshold. SPEC Web99 [73] uses such a technique, however, we believe that our simulations are a good stress test of the le system, especially as it relates to le system meta data. 7.3 Server Level Experiments The experiments described in this section were run at the server level, that is, within the le system ....

SPEC. SPECweb99 release 1.02. Technical report, Standard Performance Evaluation Corporation, July 21, 2000.

....have up to 4 processors and 16 Gbytes of memory. There is a maximum of 8 such boards in a node (cabinet) and then up to 4 nodes can be connected together via a high speed crossbar. The system has nearly uniform memory access across it potential 512 GBytes of memory. As the SPEC OpenMP benchmarks, [7], show, it is possible to obtain parallel speed ups using the full 128 processor configuration on non trivial applications. The parallel programming environment is provided by Fortran and C compilers that support OpenMP Version 1.1, 8] Both compilers also have extensive support for the automatic ....

SPEC Organization, "Standard Performance Evaluation Corporation OpenMP Benchmark Suite", June 2001.

....(ITC) cache explained in Section 3. The combination of the HBTC and the ITC cache is also evaluated. 5. 1 Simulation Environment In this evaluation, eight integer programs using the training input and five floating point programs using the test input from the SPEC95 benchmark suit are used [14], as shown in Table 1. We have used the SimpleScalar simulation tool set 2.0 [13] For each program, the tag comparison count on the following caches is measured by performing instruction level simulation: CTC (Conventional Tag Comparison cache) Tag comparison is performed on every cache ....

SPEC (Standard Performance Evaluation Corporation), URL: http://www.specbench.org/osg/cpu95. Koji Inoue was born in Fukuoka, Japan in 1971.

....memory to make non interfering spilled values occupy the same memory location when possible. We ran these algorithms on a suite of 122 routines, drawn from sources that include code from Forsythe et al. s book on numerical methods [11] the SPEC 89 benchmarks [24] and the SPEC 95 benchmarks [25]. Out of this suite, 59 routines required some amount of spill code, and it is on these 59 routines that the following numbers were generated. All the routines were subjected to extensive scalar optimization, including global value numbering, global constant propagation, global dead code ....

SPEC release 1.10, September 1995. Standards Performance Evaluation Corporation.

....compaction routine that colors spill memory to make non interfering spilled values occupy the same memory location when possible. We ran these algorithms on a suite of 122 routines, drawn from sources that include code from Forsythe et al. s book on numerical methods [11] the SPEC 89 benchmarks [24], and the SPEC 95 benchmarks [25] Out of this suite, 59 routines required some amount of spill code, and it is on these 59 routines that the following numbers were generated. All the routines were subjected to extensive scalar optimization, including global value numbering, global constant ....

SPEC release 1.2, September 1989. Standards Performance Evaluation Corporation.

....# , Howard Jay Siegel # , Anthony A. Maciejewski # , and Viktor Prasanna # Purdue University # University of Southern California School of Electrical and Computer Engineering Department of Electrical Engineering West Lafayette, IN 47907 1285 USA Los Angeles, CA 90089 2560 USA alis, kim42 ecn.purdue.edu yangyu, gundala, gertphol, prasanna halcyon.usc.edu Colorado State University # Department of Electrical and Computer Engineering Department of Computer Science Fort Collins, CO 80523 1373 USA hj, aam colostate.edu Abstract Real time applications continue to ....

....periodically, and these data streams are fed into applications. The applications process the data and send the output to other applications or to actuators (Figures 1 and 2) be the set of machines in the system. Each machine is characterized by its floating point and integer SPEC values [42] and the background load on its CPU and input output network links. The background loads on a machine are the utilizations of the CPU, input link, and output link before any applications are mapped on the system. be the set of applications that need to be mapped. All applications in ....

SPEC CPU '95 technical manual. Warrenton, VA, Aug., 1995. Standard Performance Evaluation Corporation.

....up to 4 processors and 16 Gbytes of memory. There is a maximum of 8 such boards in a node (cabinet) and then up to 4 nodes can be connected together via a high speed crossbar. The system has nearly uniform memory access across its potential 512 GBytes of memory. As the SPEC OpenMP benchmarks, [7], show, it is possible to obtain parallel speed ups using the full 128 processor configuration on non trivial applications. The parallel programming environment is provided by Fortran and C compilers that support OpenMP Version 1.1 1 . Both compilers also have extensive support for the automatic ....

SPEC Organization, "Standard Performance Evaluation Corporation OpenMP Benchmark Suite", June

....one, and a version with leaks, i.e. without the explicit assignments to null in the removeElementAt and removeAllElements methods. Then we compared the allocation integral, calculated as the area under the heap size curve. We measured the allocation integral in two programs from Spec JVM98 [15], javac, the Java compiler, and db, a benchmark simulating a database, on the original and modified Spec inputs. These are the only programs in the Spec suite using vectors. The measurements were done on a 400 MHz Intel PentiumII CPU with 128MB of memory, running Windows NT 4.0. In javac we ....

SPEC JVM98. Standard Performance Evaluation Corporation (SPEC), Fairfax, VA, 1998. Available at http://www.spec.org/osg/jvm98/.

No context found.

SPEC. Standard performance evaluation corporation, 2003. www.specbench.org/benchmarks.html.

No context found.

SPEC JVM Client98 benchmark. Standard Performance Evaluation Corporation.

No context found.

SPEC 95. Standard Performance Evaluation Corporation Benchmarks. <http://www.spec.org/osg/cpu95/CINT95/>, July 1995.

No context found.

SPEC, Standard Performance Evaluation Corporation. Available from <http://www.spec.org/ spec/>, 2001.

No context found.

SPEC Benchmarks. Standard Performance Evaluation Corporation. http://www.spec.org/, 1998.

No context found.

SPEC. SFS 3.0. Whitepaper 1.0, Standard Performance Evaluation Corporation, Warrenton, Virginia, 2001.

No context found.

SPEC. SPECweb99 release 1.02. Technical report, Standard Performance Evaluation Corporation, July 21, 2000.

No context found.

SPEC. Standard performance evaluation corporation, 2003. www.specbench.org/benchmarks.html.

No context found.

SPEC. Standard performance evaluation corporation, 2003. www.specbench.org/benchmarks.html.

No context found.

SPEC Benchmarks. Standard Performance Evaluation Corporation. http://www.spec.org/, 1998.

No context found.

SPEC release 1.2, September 1990. Standards Performance Evaluation Corporation.

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