HP Vectra XU 6/XXX HP Vectra XU 6/xxx, Guide to Optimizing performance - Page 57

Three Pipelines, Micro-ops, Multiple Execution Units

Get HP Vectra XU 6/XXX PDF manuals and user guides View all HP Vectra XU 6/XXX manuals
Add to My Manuals
Save this manual to your list of manuals

Page 57 highlights

2 Technical Reference The Intel Pentium Pro Processor Superpipeline Architecture Three Pipelines The key to the Pentium Pro processor's improved performance is its superpipelined architecture, which allows it to execute instructions more quickly. Each instruction is executed in a number of steps through a mechanism called a pipeline. In fact the Pentium Pro processor uses three pipelines, which allows it to execute more than one instruction simultaneously. Each Pentium Pro's pipeline consists of up to 14 individual steps. Pipelines have been used by previous generations of PC processors, such as the i486™ and the Pentium, to accelerate instruction execution. The reason for this is that x86 compatible instructions require a number of separate operations to complete. By using a pipeline, these processors are able to queue instructions through the individual operations so that one instruction can be completed with each clock cycle. Micro-ops The Pentium Pro processor takes the pipeline approach one stage further, by breaking the x86 instructions down into simpler operations called micro-ops. These micro-ops are then executed through the Pentium Pro's pipelines. The micro-ops are much simpler to handle than x86 instructions, and allow the Pentium Pro processor more flexibility in the way it executes them. The Pentium Pro can reorder the micro-ops within its pipelines for greater efficiency and can frequently execute multiple micro-ops in parallel. Multiple Execution Units The core step in a processor pipeline is the execution unit; this is where the real work is done. The pipeline steps before the execution units are mostly concerned with preparing the instruction for execution. The steps after the execution units involve the re-assembly of data and addresses. The Pentium Pro provides five execution units, giving it increased flexibility for instruction execution. To ensure that the pipelines are used to their greatest efficiency, the Pentium Pro must keep its instruction units busy. For this reason, the Pentium Pro implements two additional features for instruction execution: out-of-order execution; and speculative execution. English 51

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27
  • 28
  • 29
  • 30
  • 31
  • 32
  • 33
  • 34
  • 35
  • 36
  • 37
  • 38
  • 39
  • 40
  • 41
  • 42
  • 43
  • 44
  • 45
  • 46
  • 47
  • 48
  • 49
  • 50
  • 51
  • 52
  • 53
  • 54
  • 55
  • 56
  • 57
  • 58
  • 59
  • 60
  • 61
  • 62
  • 63
  • 64
  • 65
  • 66
  • 67
  • 68
  • 69
  • 70
  • 71
  • 72
  • 73
  • 74
  • 75
English
51
2
Technical Reference
The Intel Pentium Pro Processor
Superpipeline Architecture
Three Pipelines
The key to the Pentium Pro processor’s improved performance is its
superpipelined architecture, which allows it to execute instructions
more quickly. Each instruction is executed in a number of steps through
a mechanism called a pipeline. In fact the Pentium Pro processor uses
three pipelines, which allows it to execute more than one instruction
simultaneously. Each Pentium Pro’s pipeline consists of up to 14
individual steps.
Pipelines have been used by previous generations of PC processors,
such as the i486™ and the Pentium, to accelerate instruction
execution. The reason for this is that x86 compatible instructions
require a number of separate operations to complete. By using a
pipeline, these processors are able to queue instructions through the
individual operations so that one instruction can be completed with
each clock cycle.
Micro-ops
The Pentium Pro processor takes the pipeline approach one stage
further, by breaking the x86 instructions down into simpler operations
called micro-ops. These micro-ops are then executed through the
Pentium Pro’s pipelines. The micro-ops are much simpler to handle than
x86 instructions, and allow the Pentium Pro processor more flexibility
in the way it executes them. The Pentium Pro can reorder the micro-ops
within its pipelines for greater efficiency and can frequently execute
multiple micro-ops in parallel.
Multiple Execution Units
The core step in a processor pipeline is the execution unit; this is where
the real work is done. The pipeline steps before the execution units are
mostly concerned with preparing the instruction for execution. The
steps after the execution units involve the re-assembly of data and
addresses. The Pentium Pro provides five execution units, giving it
increased flexibility for instruction execution.
To ensure that the pipelines are used to their greatest efficiency, the
Pentium Pro must keep its instruction units busy. For this reason, the
Pentium Pro implements two additional features for instruction
execution: out-of-order execution; and speculative execution.