Thanks to visit codestin.com
Credit goes to www.scribd.com

Scribd
Upload
113 views2 pages

To Do

This document provides instructions for an interview project that involves using the SimpleScalar simulator and SPEC2000 benchmarks to evaluate the performance impact of various cache configurations, including instruction cache size, data cache size and associativity, L2 cache size, and L1 and L2 block size. The student is asked to run the benchmarks for each configuration, compare results based on instructions per cycle, cache miss rates, and other metrics, then write a report presenting the results, analysis, and conclusions.

Uploaded by

saeed khan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
113 views2 pages

To Do

This document provides instructions for an interview project that involves using the SimpleScalar simulator and SPEC2000 benchmarks to evaluate the performance impact of various cache configurations, including instruction cache size, data cache size and associativity, L2 cache size, and L1 and L2 block size. The student is asked to run the benchmarks for each configuration, compare results based on instructions per cycle, cache miss rates, and other metrics, then write a report presenting the results, analysis, and conclusions.

Uploaded by

saeed khan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Please complete this interview project

In this interview project, you will be asked to use the SimpleScalar “sim-outorder” model and the
SPEC 2000 benchmark suite to evaluate the performance impact of a number of basic cache
configuration parameters, such as the L1 and L2 cache size, associativity, block size, and so on.

The following is a list of example configurations you can start with, you are not required to exactly
follow them, and you are free to think out and try any configuration you would like to test. Actually,
if you can try some other configurations besides the following ones, I will give you some bonus on
your project.

a) Instruction cache size (compare 8KB, 32 KB and 64KB, block size and associativity follow the
default configuration)
b) Data cache size (compare 2KB, 32KB, and 128KB, all 4-way set associative, block size follow the
default configuration )
c) Data cache associativity (compare 1-way, 2-way, 8-way, all 32KB, block size follow the default
configuration)
d) L2 cache size (compare 128KB, 512KB, and 1MB, block size and associativity follow the default
configuration)
e) L1 and L2 block size (compare 32B, 64B, 128B, and 256B, cache size and associativity follow
the default configuration)

You are suggested to run the simulator using the default parameters except for the parameters
under study to better evaluate the impact of the configurations on the performance. You are
suggested to run each SPEC2000 benchmark for 100 million instructions after skipping 100
million instructions. Note: skipping a large number of instructions is quite important to warm up the
cache, and run a relatively large number of instructions is important to get more accurate results.

The results should be compared by using the number of instructions per cycle (IPC), cache miss
rate, the number of load/store instructions, and any other useful information you can obtain from
the simulator outputs for the executed program.

You are asked to write a report in which you should present the results (a good method to present
results is drawing figures), analyze the results, and draw conclusions about the impact of the above
cache configurations on the performance.
Report format highlights:
• Report must be submitted in printable PDF format.
• Text must be in 11pt font size, double column, single-spaced.
• There is no page limitation on the report. Generally, the report should be approximately 2-4
pages long.

Simulator and Benchmarks


You are able to download the SimpleScalar simulator and the SPEC2000 benchmark suite

Whereas SimpleScalar is an older simulator, you will find that it is well established and still
widely used. Here are a few links to get you started, but a quick google search should provide you
with everything that you need:
SimpleScalar home page
SimpleScalar Technical report
SimpleScalar Hacking Guide (for version 2 but you will use the slightly different version 3)
SimpleScalar Toolset version 3.0

You might also like