• Documents
  • Authors
  • Tables
  • Log in
  • Sign up
  • MetaCart
  • DMCA
  • Donate

CiteSeerX logo

Tools

Sorted by:
Try your query at:
Semantic Scholar Scholar Academic
Google Bing DBLP
Results 1 - 10 of 16,611
Next 10 →

TABLE I Embedded Processors for Multimedia

in unknown title
by unknown authors 1998

Table 1: Comparison of two modern embedded processor

in The Filter Cache: An Energy Efficient Memory Structure
by Johnson Kin, Munish Gupta, William H. Mangione-smith 1997
Cited by 161

Table 1. Typical Embedded Processor Interrupt Performance

in
by unknown authors
"... In PAGE 3: ...ine. Response and exit times together represent overhead. Worst case performance must be considered during system design since taking the best or average case will lead to condi- tions such as buffer overflow/underflow. While performance for a particular system configuration and operating system may vary, Table1 shows that response time for an interrupt with a full context save for high performance processors is on the order of several microseconds. A major component of response and exit times is saving and restoring the state of the interrupted context.... ..."

Table 3.4: Some embedded processors.

in A Comparison of Address Translation Mechanisms for Virtually-Addressed Caches in Embedded Systems
by Harvey Tuch, University Of New South Wales, Supervisor Prof, Gernot Heiser 2002

Table 4. Embedded Processor Test Time Comparison

in Embedded Processor Based Built-In Self-Test and Diagnosis of Logic and Memory Resources in FPGAs
by Daniel Milton, Sachin Dhingra, Charles E. Stroud
"... In PAGE 6: ... Total test time during BIST of FPGAs is dominated by two factors: downloading the BIST configurations and retrieving the BIST results at the end of the BIST sequence. We use the time required for full configuration of the FPGA and full configuration memory readback as a baseline for determining the speed-up in test time summarized in Table4 . The download time can be significantly reduced by partial reconfiguration where only those frames that have configuration data changes from one BIST configuration to the next need to be written.... In PAGE 6: ... Using the multiple frame write capability in Virtex-4, there is a further reduction in the number of frames that must be written. It should be noted that the number of frames for multi-frame write in Table4 is given in terms of equivalent frames to account for the additional configuration register commands required to perform the multi-frame write operation. Using dy- namic partial reconfiguration, the contents of the ORA remain unchanged from one BIST configuration to the next such that failure indications can be accumulated in the ORA flip-flops over a sequence of BIST configurations with the results read at the end of the BIST session.... ..."

Table 1 Execution time of the modules running on embedded processor.

in Software Synthesis of SystemC Models
by Virginia Polytechnic Institute, Brijesh Sirpatil, Brijesh Sirpatil 2001
"... In PAGE 5: ...v List of tables. Table1 Execution time of the modules running on embedded processor.... ..."
Cited by 1

Table 3: Features and sizes of the workstation processor model and the embedded processor model. Parameter Workstation Model Embedded Model

in EVALUATION OF BRANCH PREDICTION ON PROCESSOR ENERGY CONSUMPTION BY
by John Thomas Strasser
"... In PAGE 9: ...---25 Table 2: Description of the modifications made to the basic Wattch simulator.--------------------25 Table3 : Features and sizes of the workstation processor model and the embedded processor model. --------------------------------------------------------------------------------------------------26 Table 4: Energy contributions for both the embedded and workstation processors.... In PAGE 34: ...3 Processor Models Two models were used in this study to distinguish between the high-performance market and the portable computing market. Table3 highlights the features of each processor model. The workstation model does not accurately represent any particular processor, but is typical of a modern, high-performance processor.... ..."

Table 4: Energy contributions for both the embedded and workstation processors.

in EVALUATION OF BRANCH PREDICTION ON PROCESSOR ENERGY CONSUMPTION BY
by John Thomas Strasser
"... In PAGE 9: ...--------------------25 Table 3: Features and sizes of the workstation processor model and the embedded processor model. --------------------------------------------------------------------------------------------------26 Table4 : Energy contributions for both the embedded and workstation processors. ---------------26 Table 5: Subset of SPEC 2000 integer benchmarks used for simulation.... In PAGE 35: ...225 GHz 1.225 GHz The components of the energy model and their contribution to processor energy consumption are listed in Table4 for both the embedded model and the workstation model. In both processors, the clock network consumes the most energy of any single component.... ..."

Table 3. The characteristics of the embedded architectures and processors we compare in this paper.

in Vector vs. Superscalar and VLIW Architectures for Embedded Multimedia Benchmarks
by Christoforos Kozyrakis, David Patterson 2002
"... In PAGE 5: ... We compared the density of the VIRAM executables to that for a set of CISC, RISC, and VLIW architectures for high performance embedded processors. Table3 presents the five alternative architectures, as well as the character- istics of the specific processors we studied. We retrieved their code size and all performance scores for the bench- marks from official EEMBC reports submitted by the cor- responding vendors [7].... ..."
Cited by 20

Table 1. Register configurations for a number of embedded and workstation processors.

in The Need for Large Register Files in Integer Codes Abstract
by Matthew Postiff, David Greene, Trevor Mudge
"... In PAGE 2: ... This lack of agreement in the architecture and compiler research communities is mirrored in commercial processor designs. Table1 shows the register configuration of several embedded and workstation-class processors. While it is true that the register set size on some of these machines was constrained by backward compatibility and cost concerns, it is interesting to note the wide variety of register configurations.... ..."
Next 10 →
Results 1 - 10 of 16,611
Powered by: Apache Solr
  • About CiteSeerX
  • Submit and Index Documents
  • Privacy Policy
  • Help
  • Data
  • Source
  • Contact Us

Developed at and hosted by The College of Information Sciences and Technology

© 2007-2019 The Pennsylvania State University