A set associative type cache memory incorporated in a microprocessor includes memory arrays arranged in a group for each line, and only a memory array selected by a line address selected at the time of operation is made operative, thus reducing power consumption.
A method is for manufacturing a cache memory device for writing cache data into and reading cache data from a storing region of a memory cell array designated according to a portion of an address signal and a selection signal from a 1/2 multiplexer. In a first step of the method, a common base of the 1/2 multiplexer applicable to both a direct-map type cache memory device and a set-associative type cache memory device is manufactured. Then, in a second step of the method, one of the direct-map type cache memory device and the set-associative type cache memory device is manufactured using the common base of the 1/2 multiplexer. The 1/2 multiplexer included in the direct-map type cache memory device outputs another portion of the address signal as the selection signal, and the 1/2 multiplexer included in the set-associative type cache memory device selectively outputs one of a write way signal and a read way signal as the selection signal according to a write enable signal.
A microprocessor is provided with an integral, two level cache memory architecture. The microprocessor includes a microprocessor core and a set associative first level cache both located on a common semiconductor die. A replacement cache, which is at least as large as approximately one half the size of the first level cache, is situated on the same semiconductor die and is coupled to the first level cache. In the event of a first level cache miss, a first level entry is discarded and stored in the replacement cache. When such a first level cache miss occurs, the replacement cache is checked to see if the desired entry is stored therein. If a replacement cache hit occurs, then the hit entry is forwarded to the first level cache and stored therein. If a cache miss occurs in both the first level cache and the replacement cache, then a main memory access is commenced to retrieve the desired entry. In that event, the desired entry retrieved from main memory is forwarded to the first level cache and stored therein. When a replacement cache entry is removed from the replacement cache by the replacement algorithm associated therewith, that entry is written back to main memory if that entry was modified. Otherwise the entry is discarded.
The invention provides a cache architecture that selectively powered-up a portion of data array in a pipelined cache architecture. A tag array is first powered-up, but the data array is not powered-up during this time, to determine whether there is a tag hit from the decoded index address comparing to the tag compare data. If there is a tag hit, during a later time, a data array is then powered-up at that time to enable a cache line which corresponds with the tag hit for placing onto a data bus. The power consumed by the tag represents a fraction of the power consumed by the data array. A significant power is conserved during the time in which the tag array is assessing whether a tag hit occurs while the data array is not powered-on at this point.
A method for accessing a cache memory which facilitates incremental and store requests off an applied base address request increases the bandwidth of cache via the use of an internal address generation facility built into the memory's decoding circuitry. The introduction of an internal address generation facility simplifies extraneous control of typical requesters built into a memory system. The method also reduces power consumed by requests which exploit the memory's internal address generation facility. Power consumption is further reduced in a set associative cache memory system by enabling one set of sense amplifiers during an incremental fetch.
A memory design which facilitates incremental and store requests off an applied base address request increases the bandwidth of cache via the use of an internal address generation facility built into the memory's decoding circuitry. The introduction of an internal address generation facility simplifies extraneous control of typical requesters built into a memory system. The memory design also reduces power consumed by requests which exploit the memory's internal address generation facility. Power consumption is further reduced in a set associative cache memory system by enabling one set of sense amplifiers during an incremental fetch.
