EasyManuals Logo

AMD K5 User Manual

AMD K5
406 pages
To Next Page IconTo Next Page
To Next Page IconTo Next Page
To Previous Page IconTo Previous Page
To Previous Page IconTo Previous Page
Page #47 background imageLoading...
Page #47 background image
Cache Organization and Management 2-23
18524C/0Nov1996 AMD-K5 Processor Technical Reference Manual
If an internal snoop hits its target, the processor does the fol-
lowing:
During Instruction-Cache Read MissThe line in the data
cache, store buffer, or writeback buffer is written back (if
modified) and invalidated, and the instruction-cache read is
performed again. If the data-cache line was modified, a
copy of the writeback data is passed directly to the instruc-
tion cache, thus avoiding a line-fill bus cycle after the write-
back bus cycle.
During Data-Cache Read MissThe line in the instruction
cache, prefetch cache, or line-fill buffer stays valid, and the
data-cache read is performed as a single, non-cacheable
read.
During Data-Cache Write MissThe line in the instruction
cache, prefetch cache, or line-fill buffer is invalidated, the
reorder buffer invalidates all instructions in the pipeline
following the instruction that initiated the snoop, and the
data-cache write is performed.
The AMD-K5 processor, like the 486 processor but unlike the
Pentium processor, requires a jump (near or far) after a self-
modifying write to clear the prefetch cache. However, both the
AMD-K5 and the Pentium processors require a serializing
instruction after self-modifying code whose physical address is
aliased to multiple linear addresses.
2.3.7 Buffers
Several buffers are associated with the instruction and data
caches, as described below.
Line-Fill Buffers The processor has two 16-byte line-fill buffers in the bus inter-
face unit, one of which is used during instruction-cache line
fills and the other during data-cache line fills. The buffer holds
half of the 32-byte burst cycle that the processor drives in
response to a cacheable fetch miss.
Instruction-cache lines are 16 bytes wide. During fetch misses,
the first 16 bytes of the burst go through the prefetch cache to
the instruction cache and/or byte queue. The remaining 16
bytes from the 32-byte burst cycle, if they are not used immedi-
ately thereafter to fill the prefetch cache, are held in a 16-byte
line-fill buffer in the bus interface unit for a possible future

Table of Contents

Questions and Answers:

Question and Answer IconNeed help?

Do you have a question about the AMD K5 and is the answer not in the manual?

AMD K5 Specifications

General IconGeneral
ManufacturerAMD
ModelK5
Architecturex86
MicroarchitectureK5
Introduction Year1996
Clock Speed75 - 133 MHz
Core Count1
SocketSocket 7
Core steppingSSA/5, 5k86
Voltage3.3V
Transistors4.3 million
L1 Cache8 KB (data) + 16 KB (instruction)
FSB50 MHz to 66 MHz
Process Technology350 nm

Related product manuals