Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Chronology for understanding computer architecture

Status
Not open for further replies.

gary36

Full Member level 4
Joined
Mar 31, 2018
Messages
208
Helped
0
Reputation
0
Reaction score
0
Trophy points
16
Activity points
1,806
I have an elementary question on computer architecture. What is the relation between von neumann ,harward to VLIW or superscalar processor. Can von nuemann have a VLIW implementation? This confuses me as there are multiple interpretations. Similarly, I want a relation between CISC, RISC vs VLIW, superscalar, vector processor?
 

Von neumann and Harvard are the two architecture types that
classify a machine in terms of memory usage.

program and data in same memory.(Von neumann)
program separate and data separate(Harvard)

All others you mentioned belong to instruction classification(CISC,RISC,VLIW...)
 

There are numerous interpretation and that confuses me. Flynn's classification is about parallel computing. Is harvard not a parallel architecture?. We can always build a SIMD machine with harvard architecture (SHARC DSP). So the real confusion is which one come first. All of them have suffix "architecture" and I am not getting to the point of defining processor specs. For eg superscalar is micro architectural principle. Now I need to know the link of superscalar with ISA and von-nuemann/harvard?
 

VonNeumann/Harvard came first.(1940/50s)
Flynns classification afterwards(60s)

SIMD can be in Harvard(as in DSP) or in VonNeuman(x86).
Harvard or Modified Harvard preferred for performance.
 

My final take on what you explained is that, first choose ISA, then the model (harvard/von neumann), then comes the microarchitecture to realize the performance. if this is right, everything makes sense, when studying difference classes of CPU.
 

My final take on what you explained is that, first choose ISA, then the model (harvard/von neumann), then comes the microarchitecture to realize the performance. if this is right, everything makes sense, when studying difference classes of CPU.

Yes. ISA (for the application in mind) first.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top