Reinhold P. Weicker. Dhrystone: a synthetic systems programming benchmark. Commun. ACM, 27(10):1013--1030, 1984. 15  Home/Search   Document Not in Database   Summary   Related Articles   Check

....Results System Call Without RaceGuard With RaceGuard Overhead Stat non existent file 4.3 microseconds 8.8 microseconds 104 Open non existent file 1.5 milliseconds 1.44 milliseconds 4 Fork 161 microseconds 183 microseconds 13 1. After the venerable Dhrystone integer performance benchmark [16], which in turn is a reference to the Whetstone floating point benchmark. Table 2: Khernelstone Macrobenchmark, in Seconds Real Time User Time System Time Without RaceGuard 10,700 8838 901 With RaceGuard 10,742 8858 904 Overhead 0.4 0.2 0.3 under Webstone. While this does not make ....

ReinholdP. Weicker. Dhrystone: A Synthetic Systems Programming Benchmark. Communications of the ACM, 27(10):1013--1030, October 1984.

....of the system (server and the NADs) grows linearly with the number of NADs when the NADs have third party transfer capability. 3.1.3. CPU Usage In this experiment, we compare the CPU usage on the server with and without SPIRAL by measuring the remaining CPU power using the dhrystone benchmark [25]. Running the benchmark on the server during idle time gives 628,930.8 dhrystones per second. We ran the benchmark on the server while the clients are retrieving large files from all the five disk machines at the same time. Table 1 lists the results reported by dhrystone. 0 10 20 30 40 50 ....

R. P. Weicker. Dhrystone: A synthetic systems programming benchmark. Communications of the ACM, 27:1013-- 1030, 1984.

....smaller caches than they are currently used in workstation processors are su#cient to achieve e#cient execution. In Chapter 4, we will see that caches in the order of 16 32kB are a good design choice. 55 There is a long history of developing synthetic benchmarks (Whetstone [CW76] Dhrystone [Wei84] and realistic benchmarks (SPEC [Sta95] The most popular benchmark associated with workstations have been the SPEC suites. Benchmarks aimed towards other application classes have also been successfully developed with two examples being TPC [Tra98] for transaction processing applications, and ....

Reinhold Weicker. Dhrystone: A synthetic systems programming benchmark. Comm. of the ACM, 27(10):1013--1030, October 1984.

....applications. The proposed dissertation investigates architectures that allow in band distribution of processing instructions and that can support more complex processing. 5. 3 Benchmarks There is a long history of developing benchmark programs of both the synthetic (Whetstone [11] Drystone [48]) and real (SPEC [41] varieties. While the most popular benchmark associated with workstations has been the SPEC integer and floating point suite, benchmarks aimed towards other application classes have also been developed and used with some success. Two examples are TPC [45] a benchmark ....

R. Weicker. Dhrystone: A synthetic systems programming benchmark. Comm. of the ACM, 27(10):1013--1030, Oct. 1984.

....These features are important for some of the emerging applications targeted at cell phones such as games, video and audio players. The availability of appropriate benchmark suites has spurred innovation in hardware and software systems in the past. For example, benchmark suites such as Dhrystone[27], Linpack[11] SPECJVM[5] SPECjbb[4] SPEC2000[3] Mediabench[15] MiBench[13] PowerStone[16] DSPstone[30] Perfect Club[6] have been used widely for evaluating different architectures and software optimizations. These benchmark suites are generally targeted at non Java environments or ....

R. Weicker. Dhrystone: A synthetic systems programming benchmark. Communications of the ACM, 27:1013--1030, 1984.

....execution time to be consumed in the internal transition function, which allows us to simulate code to be executed. 5. #extdhrystones: indicates the execution time to be consumed in the external transition function, which allows us to simulate code to be executed. We used the Dhrystone benchmark [WEI84] to generate different workload in both internal and external transition functions. Dhrystone code is a synthetic benchmark intended to be representative for system (integer) programming. The #intdhrystones and #extdhrystones parameters allow us to execute timeconsuming code inside the internal ....

Weicker, R. "Dhrystone: A synthetic systems programming benchmark". In Communications of the ACM, volume 27, pages 1013--1030, 1984.

....to these low level metrics does not necessarily show better performance in real life applications. 4.1.2 High Level Benchmarks 4.1.2.1 Synthetic Benchmarks Another approach to the measurement of the computing performance is synthetic benchmarks. These include Whetstone and Dhrystone benchmarks [7, 35]. In these benchmarks, many computation modules are included based on the frequency of each module in sampled applications to represent real applications. However, since these benchmarks are synthetic, in many aspects, there are differences between them and the real applications, such as ....

R. P. Weicker, "Dhrystone: A synthetic systems programming benchmark," Comm. ACM,Vol. 27, No. 10, p.1013-1030, 1984.

....of the system (server and the disks) grows linearly with the number of disks when the disks have third party transfer capability. 3. 3 CPU Usage In this experiment, we compare the CPU usage on the server with and without SPIRAL by measuring the remaining CPU power using the dhrystone benchmark [24]. Running the benchmark on the server during idle time gives 628,930.8 dhrystones per second. We ran the benchmark on the server while the clients are retrieving large files from all the five disks at the same time. Table 1 lists the results reported by dhrystone. For fair comparison, the SPIRAL ....

R. P. Weicker. Dhrystone: A synthetic systems programming benchmark. Communications of the ACM, 27:1013-- 1030, 1984.

....machine. The amount of time consumed during execution would then depend on the ratio of the speed of the execution machine to the speed of the reference machine. However, We have chosen to let the user supply the speed of the reference machine in MIPS as measured by the Dhrystone benchmark [12]. The duration of the computation now depends on the ratio of the speed of the execution machine to the speed of the chosen reference machine. The computation within the template job is an adapted version of the Dhrystone benchmark. This benchmark consists of a set of functions that are executed ....

Reinhold P. Weicker. Dhrystone: A Synthetic Systems Programming Benchmark. Communications of the ACM, 27(10):1013--1030, October

....are likely to appear soon [2] 10] 21] A key design issue is just how to select from the numerous alternatives given the available chip area and performance implications of each decision. There is a long history of developing benchmark programs of both the synthetic (Whetstone [8] Dhrystone [30]) and real (SPEC [24] varieties. While the most popular benchmark associated with workstations have been the SPEC suites, benchmarks aimed towards other application classes have also been successfully developed with two examples being TPC [27] for transaction processing applications, and SPLASH ....

Weicker, R. [1984]. "Dhrystone: A Synthetic Systems Programming Benchmark," Comm. of the ACM. 27:10, 10131030.

....execution of software that performs control system functions (e.g, monitoring and evaluating sensor data, or controlling an actuator) Usability requirements dictate that the benchmark application components be dynamically controllable, relocatable and replicable. Traditional computing benchmarks [14], 8] 19] 15] are inadequate for characterizing dynamic distributed real time systems. Primarily, they were not designed to exhibit behavior characteristic of control systems. Also, they focus on component level benchmarking, and thus they do not SUN Workstation SUN Workstation Global ....

R.P.Weicker. Dhrystone: A synthetic systems programming benchmark. In Communications of the ACM, volume 27, pages 1013--1030, 1984.

....4 Table 2) Naturally, this does not include the cost of dMT s extra multiplexors, nor dMT s potential savings in bypassing hardware. 4. 2 Benchmarks For our experiments, we used a mix of benchmark programs from the SPEC95 [19] and MediaBench [11] suites, and we also used the Dhrystone benchmark [22]. Rather than show data averaged across all benchmarks as in [6] we chose pairs of benchmarks to run together. This lets us show a richer variety of reactions to BMT and dMT; some benchmarks have very distinctive behaviors. Naturally, it was impossible to present data for all possible pairs, so ....

R. P. Weicker. Dhrystone: A synthetic systems programming benchmark. Communications of the ACM, Oct. 1994. 10

....The program itself was constructed using 11 modules containing loops whose iteration counts were treated as variables; the values of these variables were then selected to allow the frequencies in the synthetic program to match the observed frequencies. A third example is the Dhrystone benchmark (Weicker 1984, 1988) Whereas Whetstone primarily measured floating point performance, Dhrystone was characterized by its author as a systems programming benchmark. To determine the frequency of individual high level language constructs, 16 different data collections drawn from several languages were used. ....

Weicker, R. P. (1984). "Dhrystone: a synthetic systems programming benchmark," Communications of the ACM 27(10), 1013-30.

....direct symbolic execution method presented here, and the authors focus on more ideal machine models. However, extremely accurate performance estimates are achieved. In some respects the goals of our approach are similar to the goals of the creators of the synthetic Whetstone [CWW76] and Dhrystone [Wei94] benchmarks. However, we attempt to provide an automated approach to generating a synthetic benchmark. Furthermore this automated approach is based upon analysis of real programs and is demonstrably more representative of actual machine performance. 7 Future work This work can be extended in ....

R. P. Weicker. Dhrystone: A synthetic systems programming benchmark. Comm. ACM, October 1994.

No context found.

Reinhold P. Weicker. Dhrystone: a synthetic systems programming benchmark. Commun. ACM, 27(10):1013--1030, 1984. 15

No context found.

R. P. Weicker. Dhrystone: A Synthetic Systems Programming Benchmark. Communications of the ACM, 27(10):1013-1030, October 1984.

No context found.

R.P. Weicker, "Dhrystone: A Synthetic Systems Programming Benchmark," Comm. of the ACM, vol. 27, no. 10, Oct. 1984.

No context found.

R. P. Weicker. Dhrystone: A synthetic systems programming benchmark. Communications of the ACM, Oct. 1994.

No context found.

Reinhold P.Weicker. Dhrystone: A synthetic systems programming benchmark. Communications of the ACM, 27#10#:1013#1030, October 1984.

No context found.

R.P.Weicker,"Dhrystone: A synthetic systems programming benchmark," CACM, 27 (10), pp. 1013-1030, 1984.

No context found.

R. Weicker. DHRYSTONE : A Synthetic Systems Programming Benchmark. Communications of the ACM , 27(10):1013--1030, October 1984.

No context found.

R. P. Weicker, "Dhrystone: A Synthetic Systems Programming Benchmark," Commun. ACM 27, No. 10, 1013--1030 (1984).

No context found.

R.P. Weicker. Dhrystone: A Synthetic Systems Programming Benchmark. Communications of the ACM, 27(10):1013-1030, October 1984. 12

No context found.

R. Weicker. Dhrystone: A Synthetic Systems Programming Benchmark. Communications of the ACM, 27(19):2013--1030, October 1984.

No context found.

Reinhold P. Weicker, "Dhrystone: A Synthetic Systems Programming Benchmark," Communications of the ACM, October 1984.

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