WHITE PAPER (cont’d) Compaq Workstations - Key Technologies
- 6 -
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Diagram 4: Multiprocessing Support
PROCESSOR TECHNOLOGIES
The Compaq Professional Workstation 6000 uses Intel’s Pentium II processor, which incorporate
Dynamic Execution and the Dual Independent Bus architecture technologies to provide the
highest performance next generation processors.
Dynamic Execution
The Pentium II processor design is based on Intel’s Dynamic Execution technology that
combines three advanced processing techniques to increase the performance. The three
techniques are multiple branch prediction, dataflow analysis and speculative execution.
Multiple branch prediction means that the processor looks ahead several steps in the software
and predicts which groups of instructions are likely to be processed next. This increases the
amount of work delivered to the processor, improving how efficiently the processor is used and
ultimately, system performance.
Dataflow analysis analyzes which instructions are dependent on each other's results, or data.
This technique allows an optimized schedule of instructions to be created so they can be
processed in the most efficient order.
Speculative Execution means that the instructions are then carried out based on the schedule
created by the dataflow analysis. Thus, the entire processor's superscalar processing power is
kept busy, boosting overall software performance.
This combination of new technology enables the Pentium II processor to deliver the processing
performance workstation applications requires.
Dual Independent Bus Architecture
The Pentium II uses Intel’s Dual Independent Bus architecture providing two (dual) independent
buses versus the single bus architecture of the Pentium processor. The dual buses are the L2-
cache-to-processor bus and the processor-to-main-memory system bus. This design delivers up to
three times the bandwidth of a single bus architecture processor because the buses can work
independently which essentially doubles the throughput. This technology will enable the
evolution of today’s 66MHz system memory bus to the faster 100MHz system memory bus later
this year.
To Next Page
Loading...
Need help?
General