1A Pentium III Branch Cost What is slowdown as compared to the ideal pipeline fo

Question

1A Pentium III Branch Cost

What is slowdown as compared to the ideal pipeline for the Pentium III?

1B Pentium 4 prediction rate

Given this data, what was the branch prediction rate for the Pentium 4?

1C Pentium 4 Branch Cost

What is the slowdown as compared the ideal pipeline for the Pentium 4?

1D Ideal Speedup

Considering only the improvement in clock speed (and not the effects of branch prediction or any other improvements), what is the ideal speedup expected for the Pentium 4 over the Pentium III? Note that the difference in pipeline depths is already reflected in the change in clock speeds, so should not appear in your answer.

1E Speedup with Branch Prediction

What is the overall speedup considering both the change in clock rate and the effect of branch prediction and branch penalties (but not considering other architectural improvements) of the Pentium 4 as compared to the Pentium III?

Explanation / Answer

In terms of Computer Architecture, A branch is an instruction that causes the computer to start executing another instruction sequence from its default execution sequence.These are at times termed as Jump Instructions.

1 A ) Pentium III Branch Cost:

This can be estimated using the following terms

i. Branch Type - 2-level smith

ii Branch History Table(BHT) -

Constructive: Prediction is incorrect but Predicted correctly

Desctructive: Prediction is correct but predicted incorrectly

The total entries for this is 4096

iii Branch Target Buffer(BHB)

This helps in decoding the address of the next instruction to be executed.

total 512 entries

iv Branch Prediction

Dynamic Branch Prediction for Pentium III

Decisions for the Predictions may change during runtime or program execution

v Return Address Stack

Pushing the instruction address onto the stack after the Procedure call

Total 4/6 entries

vi Branch Penalties

7-15 cycles

1 B) PENTIUM III PREDICTION RATE

Branch Prediction is used when there are more instructional stages to be executed.This concept was introduced ever since the concept of pipelining was brought in. Whenever the Processor encounters an 'if-else' statement, it will use branch prediction to avoid the wait time.

Predicting which branch has to selected and which is not

Predicting the destination address of the branch

The Prediction rate or accuracy of a Pentium III on an average is around 90%.

1 C ) PENTIUM IV BRANCH COST

i Branch History Table(BHT)

4K Entries

ii. Branch Prediction

Dynamic Branch Prediction for Pentium IV /Static Branch Prediction

iii. Prediction Rate of Pentium IV

It is 93 to 94% higher than Pentium III

IV. Return Address Stack

16 entries

V) IDEAL SPEEDUP

Speedup:

Performance measure of a system while execution.

Increasing the speed of a system or the Processor

Clock rates

PENTIUM III 1000 MHz

PENTIUM IV 2000 MHz

Data Bus in Bits

PENTIUM III 32 bit

PENTIUM 32/64 Bit

Core of the Processor

PENTIUM III / IV Single Core

Speedup

Both has a similar Speedup

0.92 when a single core is used

2.94 for all 3 cores

This speedup is measured using Amdahl's law

Related Questions

Navigate

• Browse (All)
• Browse 1
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.