Module 5 Pentium Processor
Module 5 Pentium Processor
Pentium Processor
➢Wider Data Bus
➢Dedicated Instruction Cache
➢Dedicated Data Cache
➢Parallel Integer Execution
➢Floating Point Unit
➢Branch Prediction
➢New Instructions
➢Speed
Pentium performs integer instructions in a five-stage pipeline.
➢PF - Prefetching
➢D1 – Instruction Decoding
➢D2 – Address Generation
➢EX – Execution
➢WB – Write Back
➢Prefetch
➢Instruction Decode 1
➢Instruction Decode 2
➢Execution
➢FP Execution 1
➢FP Execution 2
➢Write FP Result
➢Error Reporting
Two basics cache architectures are –
1. Look-through cache design
2. Look-aside cache design
1. Reduces the system and memory bus utilization.
2. Allows bus concurrency.
3. Complete write operations in zero wait states using posted writes.
1. Cache miss cycles complete faster
2. Simplicity of designs
3. Implementation cost is lower
➢Write-through Cache Designs
➢Buffered or Posted Write-through Designs
➢Write-back Cache Designs
Pipeline Stages NA 2 3 3 5 5