Advances in Computers: Parallel, Distributed, and Pervasive ComputingMarvin Zelkowitz, Suyel Namasudra Elsevier, 2005. gada 19. maijs - 312 lappuses The term computation gap has been defined as the difference between the computational power demanded by the application domain and the computational power of the underlying computer platform. Traditionally, closing the computation gap has been one of the major and fundamental tasks of computer architects. However, as technology advances and computers become more pervasive in the society, the domain of computer architecture has been extended. The scope of research in the computer architecture is no longer restricted to the computer hardware and organization issues. A wide spectrum of topics ranging from algorithm design to power management is becoming part of the computer architecture. Based on the aforementioned trend and to reflect recent research efforts, attempts were made to select a collection of articles that covers different aspects of contemporary computer architecture design. This volume of the Advances in Computers contains six chapters on different aspects of computer architecture. Key features:
|
No grāmatas satura
1.–5. rezultāts no 76.
... dynamically issues multiple instructions and VLIW statically issues multiple instructions at each cycle. These techniques are pioneers (and now the basis) of modern computer architecture designs. In this chapter, we describe and ...
... Dynamic Dynamic Static Difficulty of Design Relatively easy compared to the other two. Complex design issues Run-time tracking. Complex design issues Compiler support. Keywords Higher MHz, latch delay, clock skew. Dynamic scheduling ...
... dynamic (using branch target address cache and branch history) and static (using the direction of the branches) ... dynamically predict those branches (to be taken), and (3) provide branch folding of the predicted branches if possible. By ...
... dynamic scheduling which uses hardware to schedule the instructions at run-time. On the contrary, VLIW architectures depend on static scheduling which schedules the instructions at compile time in a static way [11]. We will discuss the ...
... dynamic scheduling of the Superscalar model is strongly based on the early Tomasulo [31] or Scoreboard algorithms [30]. 4.1.2. Dynamic. Scheduling. of. Superscalar. Superscalar architectures achieve dynamic instruction scheduling by using a ...
Saturs
1 | |
35 | |
Characterizing Resource Allocation Heuristics for Heterogeneous Computing Systems | 91 |
Power Analysis and Optimization Techniques for Energy Efficient Computer Systems | 129 |
Flexible and Adaptive Services in Pervasive Computing | 165 |
Search and Retrieval of Compressed Text | 207 |
back matter | 263 |
index | 273 |
Contents of Volumes in This Series | 285 |