Previous | Next --- Slide 23 of 36
Back to Lecture Thumbnails
kayvonf

Write-through caches ensure memory always has the most up to date value stored in an address. This simplifies coherence, since to obtain the latest value stores at an address, the requesting processor can simply issue a memory load. Write-back caches pose new challenges to maintaining coherence, since the most recent value stored at an address may be present in a remote processor's cache, not in memory.

jmnash

It seems as if there is a trade-off here between memory access time and processor communication time. When using a write-through cache, as Kayvon mentions above, it is not only simple to maintain cache coherence, but there also does not need to be any communication between processors. However, every time there is a write, the processor has to communicate with memory. With a write-back cache, processors must communicate in some way, direct or indirect, to obtain the correct values of the data they're using. However, this will spend less time going back to memory. So which method is faster might depend on the computer, but it seems like write-back will probably be faster most of the time.