I can't seem to figure this out from the related wikipedia pages:
Add a comment
|
5 Answers
See Wikipedia's Modified Harvard Architecture page on Harvard or von Neumann?:
Three characteristics of Harvard architecture machines may be used to distinguish them from von Neumann machines:
Instruction and data memories occupy different address spaces. [...]
Instruction and data memories have separate hardware pathways to the central processing unit (CPU). [...]
Instruction and data memories are implemented in different ways. [...]
The last point is because of the original Harvard machine (stored instructions on paper tape, data in electromechanical counters), but is applicable to modern pure-Harvard chips like the PIC microcontroller, which use non-volatile memory to store instructions but volatile memory for RAM.
Now, given those distinguishing characteristics, how would you categorize the MIPS?
answered Dec 11, 2009 at 18:48
Modern processors are based on the Von Neumann architecture, including x86, x64, and MIPS. As far as I know, the Harvard architecture is not used in mainstream CPUs.
-
2later x86 chips and x64 are Von Neumann/Harvard hybrids; the Harvard influence is seen in the L1 cache design (separate data/instruction caches) and other features. see en.wikipedia.org/wiki/Modified_Harvard_architecture and en.wikipedia.org/wiki/… Dec 11, 2009 at 18:37
-
Split L1D / L1I caches are nearly universal for performance reasons (a unified L1 would have to be bigger and have more read ports to have similar performance, which would make it hard to keep latency low), but that's an implementation detail. On x86, L1I is even coherent with data caches (unlike most modern RISC architectures where self-modifying code requires a special instruction to sync i-cache + pipeline with data caches). Dec 26, 2017 at 8:30
It's a Von Neuman architecture because of no separation between data and code memory. The intel 8051 is a Harvard and there is a separate memory for data and code and to fodepush ^ this IS mainstream.
as I study before MIPS structure in computer architecture , MIPs based on Von Neumann
The main difference between Harvard and Von-Neuman architecture is that of the memory. Harvard contains two separate memory: Program Memory(contains instruction set etc.) and Data Memory(containing data, operands etc.) whereas Von Neuman has only one memory(containing both Program memory and Data memory into one).
Now, as Mips has two separate memories: Program Memory, which is used for fetching the instruction from it and data memory/main memory which is used at Memory stage to store/load data. Therefore MIPS is more close to Harvard Architecture.
-
Are you talking about an embedded MIPS implementation with code in ROM or flash, but with RAM for data? Code and data still don't have separate address spaces, though, do they? Dec 26, 2017 at 8:32