Scribd
Upload
5 views2 pages

QUIZ

The document discusses the advantages of Harvard architecture, which allows for parallel processing and simultaneous access to data and instructions, enhancing execution speed. It highlights the benefits of SIMD and MIMD features in improving data processing efficiency while also addressing their respective challenges. Additionally, it covers instruction level processing techniques like pipelining and superscalar execution, which significantly boost processor performance by increasing throughput and reducing latency.

Uploaded by

kiplangatbrian007
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
5 views2 pages

QUIZ

The document discusses the advantages of Harvard architecture, which allows for parallel processing and simultaneous access to data and instructions, enhancing execution speed. It highlights the benefits of SIMD and MIMD features in improving data processing efficiency while also addressing their respective challenges. Additionally, it covers instruction level processing techniques like pipelining and superscalar execution, which significantly boost processor performance by increasing throughput and reducing latency.

Uploaded by

kiplangatbrian007
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

Harvard architecture that integrated the following:

Allowed parallel processing units tis allowed multiple execution of instructions

Allowed simultaneous access of data and instruction increasing the execution speed

Also SIMD and MIMD features were supported that included it’s capabilities in terms of efficiency in
data processing due to their respective abilities

Memory for instruction and data were placed separately and this allowed the CPU access to varios
needed instruction fast

The justifications are:

Bottle neck was reduced significantly due to the separation of memory for both data and
instruction

Increased parallelism from the integration of SIMD and MIMD that supported parallelism

Access of resources by the CPU was fast due to the separation of memory for both data and
instruction

Benefits of SIMD and MIMD

SIMD:

Lower over head

Less resources were needed since the instruction were all similar

Efficincy in data processing especially the ones that the same operation or process was needed

MIMD:

Flexibility-due to the ability of processer to act on different data needing different instruction to be
used

It was effective in tasks that tend to be independent in nature like distributed systems

Challenges

SIMD

Lack flexibility on data processing that require more than one instruction to be used on

Data alignment and synching was hard

MIMD

High over head since extra effort is needed on processing the data that require different instructions
to be used

Data alignment and synching was hard


Instruction level processing

Pipelining: Process several instruction in parallel on each stage, such as fetch, decode, execution;
this increases throughput.

Superscalar Execution: More than one instruction is issued and executed by different execution
units inside the CPU simultaneously.

Impact on Processor Performance:

More Throughput: ILP increases the processor's speed and throughput significantly by executing
more than one instruction parallel.

Low Latency: ILP decreases the latency because the processor will fully utilize the execution units
which would sit idle otherwise.

Performance Gain: High ILP can be exploited by modern CPUs to handle complex tasks faster.
However, this is actually dependent on the code's nature. For instance, code with high branching
frequency may offset the efficiency of ILP.

You might also like