Any EE or CS majors out there remember how to calculate this crap and wouldnt mind giving me a brief explanation of how its done? Something with it is just not clicking for me:

I have an application whose memory access pattern is a stream and its entire data set is 128kB. The data cache in your machine has a capacity of 64kB. It is word-addressable and direct-mapped with 32B lines. It is able to fetch one line at a time.

Given the access pattern: 0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, …, where each access is a 4B word.
a. What is the miss rate? (4)

0

Probably 0 if the machine has predictive features. Maybe 1, if it doesn't and has to load the other 64k of data?

I don't remember learning any of this stuff.

Since you're fetching 32B at a time and reading every 4B consecutively, you only need to look for a new line every 32/4 = 8 accesses. Reading address 0 will give you data from address 0 - address 31, which contains the first 8 access in their entirety. However, like Xin said, predictive features should detect that kind of access pattern ~100% of the time which gives you a miss rate of 1/(128k/4) (first one is never free in CS ;))

Oh, now I remember why I switched out of CS... despite having a 4.0 GPS at the time, this stuff is boring! Blah...

Now coding a MUD, that was amusing...

bonus

This is the last architecture class I have to take and the semester is 3/4 over. Definitely my least favorite mandatory courses. Back to software engineering next semester :P

Also thanks Erian, that helped quite a bit. I'm not sure why books and professors are incapable of explaining things without obfuscation.

Another reason is probably because you call it a 'GPS'...

LOL... oh well, typo.

Good thing I wasn't an English major.

GPS could work.. Grade Point System

... based on your GPA :-P

*head explodes/bonus*

2 words

Job Security

they want you to fail the course so they can be justified in calling you an idiot. then you have to pay them again to teach you the same exact thing that they failed to teach you the first time.

Actually I think it's more that American textbook writers get paid per page in the book which explains why they get obfuscated instead of keeping things simple. And if it's anything like in Sweden, local teachers will often write the curriculum for advanced topics in universities which means that their writing will influence their teach most likely ;)

It hilarious comparing many European and American textbooks for science. American books are almost always 2-3x thicker for the same level of course (and no, contrary to popular belief bigger isn't always better).

weird. i thought that writers were paid royalties. they don't get a percentage of the sale?

we2342rewfsdfsdaf