Friday, February 25, 2005

i dug through boxes to find an electronics part this evening and came across my textbooks from community college. fundamentals of financial accounting. college algebra. seeing them made me feel humble rather than humiliated; i am just an average guy. i used to feel that i was too good for those books. now i realize that the books are a part of me. some people go through life at a faster pace, building brilliant libraries. the nausea of life is that it is always possible to speed up, yet at the same time the amount of things we've done to define ourselves is always increasing. the moment screams "read the best things, go to the best places, be with the best people," but the sad truth is that most of us cannot handle the best. we wait in loops to find or be ready for "the best," and then we die.

now the books i'm reading are a little more challenging at least. studying computers and engineering is addictive because the battles can always be refought. being baffled today only sweetens the pot of being right tomorrow. the problem is today.

at ECE the scarce resource is time. those who make the best use of it succeed and gloat in inverse proportion to those who fail and cower.

enough prose.

Thursday, February 24, 2005

my life has become nonstop school. i feel as though i spend almost as much time carrying books around as I do reading them. slice the pie chart up and the largest wedge by far would be staring into the computer screen. not necessarily learning. in the case of my VHDL course, lining up the edges of metal traces and dealing with a crappy CAD interface. a lot of time sitting in lecture. fascinating professors this semester. two of them are on the subject of timing, making sure that read data required starts and ends within read data available. i noticed somthing that binds together many fields of electrical engineering and computer science: parallelism. the computer program runs between samples, or processes between interrups. or on a finer level, multiple steps of computer logic execute between the miniscule edges of the system clock. i wrote in my notes that "it's all the same problem." controlling execution, which entails controlling time. the fields of electromagnetism, optics, seem quaint in comparison to the herd of geeks devoted to the pursuit of speed, moore's law, capturing higher and higher frequencies. 60 jigahurtz! my digital communications professor bellowed, the new frontier, ubiquitous broadband content. speed of course. very tiny time delay. robin tsang, my embedded systems TA remarked that intel chips are now so fast that execution cycles can no longer be traced with rectangular edges so much as humps measured in nanoseconds.