Many of you might be wondering what it is exactly that I'm doing here in grad school, now that I have finished most of my course-work and have actually started working on my thesis... Well, I'd like to know that too! So if you find out please let me know! :)
All kidding aside, I am working in the area of Computer Architecture on Hardware and Software Transactional Memory. A lot of big words, especially that "transactional" one sounds kinda scary. So what does this mean?
Well, in this day and age, multi-processor systems (computers with more than one CPU) are becoming increasingly popular, especially as the limits of Moore's law are approached. The problem with more than one processor working on the same thing is similar to having more than one person assigned to the same task - how do you coordinate the use of shared resources between the two?
Locks have traditionally been the method of doing such control. Simply give exclusive access to one processor of the required resource (memory location for example), until its done. The thing is, locks severely hinder performance, and are most probably not needed all the time since there is no real conflict (will not explain why in this post). So better methods are required. This is where my area of research lies!
I have a really cool supervisor, Prof. Jim Goodman from the University of Wisconsin-Madison, my own cubicle, and all the coffee I could handle!
So, where exactly am I on the PhD path? I am done with the first year, so I am at "Advisor lures with visions of greatness" point (see graph below - taken from PhD Comics)
It's gonna be a bumpy ride - I do not expect it to be easy. But at least I will be doing something I enjoy.
It's kind of funny though, here I am, back to school, starting a journey with about 3-4 years to go. While my friends are all getting settled and stuff! Well, this is what I asked for and this is what I'm gonna get!
Anyway, I better stop slacking around and go do some research...