Effective Access Time:
Effective Memory Access Time is used in the context of paging usually. In the case that the page is found in the TLB (TLB hit) the total access time is:
TLB_hit_time = TLB_search_time + memory_access_time
In the case that the page is not found in the TLB (TLB miss) the total access time is given by sum of time to search the TLB (to find that it is not in TLB) and the time to access memory to get the page frame mapping plus the time to access memory to get the actual data, so:
TLB_miss_time = TLB_search_time + memory_access_time + memory_access_time
Effective Access Time is simply the weighted average of the above two times. That is,
Effective Access Time = TLB_miss_time * (1- hit_ratio) + TLB_hit_time * hit_ratio
Average Memory Access Time (AMAT):
Average Memory Access Time is used in the context of cache memory.
AMAT = Hit time + Miss Rate * Miss Penalty.
Cache will always be accessed, irrespective of a hit/miss whereas a TLB access time is to be considered only when a hit is there. Otherwise, in case of a TLB miss, the TLB search time need not be considered as it is negligible.
Memory Access Time:
The time required to access instructions and data in memory. This depends on design of memory hierarchy, hit rate, miss penalty and number of cache levels.
Effective Instruction Execution Time:
Instruction execution time = CPU time + Memory access time
CPU time is the time for CPU burst. i.e., the time actually spent inside the processor for doing useful computation.
Memory access time = TLB_miss_time * (1- hit_ratio) + TLB_hit_time * hit_ratio
