Smt. Kamala & Sri. Venkappa M.

Agadi
College of Engineering & Technology Lakshmeshwar – 582116
Department of Computer Science & Engineering

Subject: Advanced Computer Architecture - Sub Code: 17CS62

Topic: 1.1.3 Evolution of Computer Architecture

Mr. Vikram Shirol

Associate Professor
Dept of CSE
vikramshirol@gmail.com

Website: http://www.agadiengcollege.com : http://facebook.com/sksvlxr

: https://twitter.com/SKSVMACET_LXR
Architectural
Evolution
Lookahead, Parallelism & Pipelining:

• Lookahead techniques were introduced to prefetch instructions in

order to overlap instruction fetch/decode and execution operations
and to enable functional parallelism.

• Pipelined instruction execution includes, pipelined arithmetic

computations, and memory-access operations.
Pipelining Example
Flynn’s Classification
1. Single Instruction Stream over a Single Data Stream (SISD)
2. Single Instruction Stream over a Multiple Data Stream (SIMD)
3. Multiple Instruction Stream over a Multiple Data Stream (MIMD)
4. Multiple Instruction Stream over a Single Data Stream (MISD)
Parallel / Vector Computer
Intrinsic parallel computers are those that execute programs in MIMD
mode.
There ate two major classes of parallel computers, namely, Shared
Memory Multi Processors & Message Passing Multi Processors.
The processors in a multiprocessor system communicate with each
other through shared variables in a common memory.
Each computer node in a multicomputer system has a local memory,
unshared with other nodes. lnterprocessor communication is done
through message passing among the nodes.
A vector processor is equipped with multiple vector pipelines that
can be concurrently used under hardware control.
There are two families of pipelined vector processors:
• Memory-to-Memory architecture supports the pipelined flow of
vector operands directly from the memory to pipelines and then back
to the memory.

• Register-to-Register architecture uses vector registers to interface

between the memory and functional pipelines.
Development Layer
• A layered development of parallel computers is illustrated in the
following Figure.
Hardware configurations differ from machine to machine, even those
of the same model.

The address space of a processor in a computer system varies among

different architectures, which depends on Memory organization.

Application programs and Programming environments are machine-

independent means the user programs can be ported to many
computers with minimum conversion costs.