Tuesday, November 24, 2009

Could be a good deal...

Yesterday I got an email regarding the Department of Energy computational science graduate fellowship. Normally I'd probably ignore such an email, but for whatever reason I took a gander. Inside I found some words (as one might expect), but these were the ones in particular that caught my eye:

Benefits of the Fellowship:

  • $32,400 yearly stipend
  • Payment of all tuition and fees
  • Workstation purchase assistance
  • Yearly conferences
  • $1,000 yearly academic allowance
  • 12-week research practicum
  • Renewable up to four years
Sounds pretty decent I thought, but what's the catch? This kind of dough doesn't usually come without some strings (or steel cables) attached, so what is it, a lifetime of indentured servitude? Well the conferences are required, but they're all expenses paid on top of extra stipend for attendance, so it's more like a mandatory paid vacation. Same for the research practicum, in which you are required to use massive DoE supercomputers for whatever you want. Notice that when they say "workstation purchase assistance," they mean that they will only match the money you put up for whatever high performance computer you want. In addition, whatever school you attend has to agree to not have you working as a TA or research assistant for more than one semester. Finally, the only non-academic requirement is that you agree to consider job offers from the DoE or contractors.

As far as I'm concerned, this whole program is the best idea anyone has ever had! The thought of being paid to go to grad school makes me very, very happy. I managed to find the applicant statistics for last year and it turns out that about 1/20 people who applied got in. Assuming equal probability those odds aren't bad at all, however that's probably not a fair assumption to make; with benefits like these, it's easy to be motivated to do better in school so that my probability of selection might improve, hence the current time and my working on homework (well, ok, blogging, but motivated or not everyone needs a break now and again). I'm actually fairly confident in my grad school prospects, mainly because of my undergraduate research. This is my second semester of such, and apparently my research advisor likes me enough to propose advising me next semester as well despite him being on sabbatical (so that we'll be prepared to "hit the ground running in the summer"). I can't express how grateful I am to have found such a good fit and generally exceptional person to work with... though that statement does do a pretty good job of at least indicating the magnitude of my gratitude. As it stands, it seems that I will be graduating with 6 semesters of research experience, which, combined with being the student administrator for the CS department's linux server and a double major, ought to more than make up for some of my less than optimal grades. Nonetheless, better grades certainly aren't going to hurt, so back to the books!

Friday, November 20, 2009

HD isn't always HD

For years now HD has been a magic word, and for just as long I've found humor in its use, when not shaking my head at the naievete involved. Everybody knows that you have two options with HD, 720 or 1080. Clearly 1080 is the better option, because it's a bigger number... right? Well, yes and no. These numbers represent the vertical resolution or number of rows of pixels from top to bottom of the screen. Of course vertical resolution is only half the picture, for whatever reason the horizontal resolution is implicit: 720 has a full resolution of 1280x720, 1080 has 1920x1080. In terms of resolution, yes, 1080 is better, but this is a really restricted and possibly misleading analysis. In terms of actual clarity, a vastly more important measure is PPI (pixels per inch). Imagine for instance that the big man on the block has a 60" 1080 HD lcd screen, in his own little world he is really special for having such a ginormous TV with such crystal clarity. But in reality, his neighbor's 20" 1080 HD lcd screen looks much clearer, and the reason is simple: both TVs have the exact same number of pixels, which means that to fill the extra space the 60" has pixels that are 3 times as big (with 9 times the area), making them much easier to distinguish from the same distance, making the contrasting areas of the image look blocky and jagged. To further illustrate, imagine another neighbor has a sad little 5" 1080 HD lcd screen--in truth, he is the one to envy! The clarity of such a screen would be astounding, with 440.6 pixels per inch it could draw letters and numbers 1/100th of an inch tall, just about twice the width (diameter) of an average human hair. On the other hand, Mr. big man only has 36.7 ppi, the smallest letter his TV could draw would be 1/7th of an inch tall, close to the width (diameter) of a pencil eraser!

The reason I say HD isn't HD is that as a computer user, I'm accustomed to HHD (higher than high-def)--you probably are too, you just weren't aware of it. Suppose we ignore all this (very relevant) pixel density stuff and focus solely on resolution; the average computer monitor has been capable of resolution better than 720 for a long time. 1280x1024 is the most common computer resolution, and it has 142% the resolution of 720. It is only 63% of 1080, but 1280x1024 is rapidly going out the window--in fact, you can now get a new 22" lcd computer monitor with greater than 1080 HD resolution for $200. The discovery of this recently surprised me, that seems like a great bargain. I'm a big fan of 30" 2560x1600 monitors, but unfortunately they are tremendously expensive, so in my idle pondering and interest in value metrics I ended up deriving the very simple math to get the numbers above and a few more that relate to 30" monitors. In short, a 2048x1152 screen has only 57.6% the resolution of a 30", but can be bought for 15-20% of the price. Likewise, if you really want to match the 30" experience, a 24" 2048x1152 monitor will have the same PPI... but any smaller size with the same resolution will also have a smoother image (or higher definition) than the 30". With this perspective it's no longer a great bargain, but an amazing deal.

Wednesday, November 18, 2009

need... more... ram... MORE RAM!

Depending on who you're talking to and how much they like to debate semantics, data mining and machine learning are essentially the same thing. The important point is that incredible amounts of data are needed in order to mine golden nuggets of useful information. The Netflix dataset used for the Netflix Prize, for instance, is well over 4GB; to make matters worse, unless one writes a lot of skillfully efficient code, putting this into a data structure takes quite a bit more space. On top of that, raw data isn't very useful unless you have room to fit whatever models you're trying to construct in addition. Despite my relatively short presence and shorter sentience on this Earth, I remember a time when 4GB was a huge capacity for a hard disk. Of course, these days 4GB will fit on the increasingly outdated optical DVD format... 4GB can even fit on a plastic sliver of flash memory, less substantial than a humble dime. In a time of terabyte hard drives costing less than a trip to the grocery store, 4GB seems laughably diminutive. However, there's a significant issue here! Most people know that a computer has several types of memory: RAM and a hard drive (there are more, to be covered momentarily). Why are there two types of memory, why not just use a hard drive? The answer is simple, getting data from a hard drive takes 100,000 times longer than from ram! While a specially built computer could run with only a hard drive, it would be so unbelievably slow that nobody in their right mind would ever use it. For a good number of tasks, like listening to music and looking at pictures, a hard drive works just fine. The reason is that these things are just read--once they've been read and used, say the sound the data represents has been sent to the speakers, the data can be thrown away. However, the more important, invisible bits of data that allow a computer to run are most often handled very differently: once read and processed, the results are stored so that they might be used later. Imagine, for instance, that you have a counter (which are extremely common in computers and programming) that counts the numbers of mouse clicks. If the processer were to take the stored counter, add one, and throw the result away, then the next time the processor read the counter it would get the number that the counter started at. If you have a program that displays some message when you click 10 times, the message will never get displayed. Obviously, in order for your program to work, the cpu must be able to remember how many clicks have happened, so it must read and write. For this simple example, the time it takes to read/write from a hard drive is ok--even the fastest human clicker is inconceivably slow compared to the inner workings of a computer. However, if this count is something that is read/written millions of times a second, the time it takes to access the hard drive will be an incredible bottleneck. In fact, this kind of situation is extremely common in computers (hard drives are the biggest bottleneck in a computer), hence why we have ram; computers simply need a place that can be accessed very quickly in order to work fast enough for us not to prefer watching grass grow. For a bit of extra credit, let me point out that as far as the cpu is concerned, even ram is dreadfully slow. See why after the jump.

Thursday, November 5, 2009

The end of scrubbing tubs

As a college aged male bachelor, I know a thing or two about bathrooms that have gone uncleaned for longer than many people might think possible. To make matters worse, I use a humble castile soap that produces scum with unabashed vigor. Last week, by the combined effort of many unknowable forces, I decided to clean the bathroom, and in the process made a fantastic, incredible, revolutionary discovery (this time not involving micro fiber cloths)! I started with the typical futile effort, spraying everything with potent cleaning chemicals and scrubbing, scrubbing, scrubbing with a typical plastic bristle brush. Before long it was clear that all my effort was adding up to nothing. I decided to pull out the big guns, and began to seek a green scouring pad. I found one, and thumbed it as I saw in the same drawer a box of those new-fangled magic "eraser sponges," also known as melamine foam, I had picked up on a whim at the dollar store. Feeling my sense of adventure kick in (and not knowing what else I might ever use them for), I decided to grab a magic sponge and give it a try. The results were unbelievable. The soap scum literally rolled off every surface after just a pass or three; I had the whole shower cleaned and sparkling in under 5 minutes! Never before in my life has a shower taken less than an hour, at least, to clean, and thus my excitement for this finding. From now on I needn't fear nor need to clear a day so that I can clean the shower; maybe, just maybe, it'll get cleaned more often now... no commitment there though. I checked the box of the eraser sponge afterward, and it does actually suggest using it for cleaning the shower, which means someone, somewhere out there actually knew about this beforehand. This is hard to believe, I would have figured the news would spread like word of a Gmail outage (wildfire in this age is the new grass growing, amirite?). I don't know if they've advertised these for this purpose since my exposure to commercials is essentially nonexistent, but I'm led to believe that even if they did it would pass unnoticed and the reason is simple: from memory, bathroom cleaning products are depicted in an ineffective way. I remember them showing what looked like an evenly disgusting tile wall, oddly aesthetic in the precision of its filth, which becomes pristine after just one pass of a sponge. Yeah, sure, everyone believes that. Even with the magic sponge there's some work involved, but I think 30 seconds of someone cleaning a real bathroom in real time with real results would be an amazing commercial. It'd say "Hey, look, this actually works. We're not trying to trick you using magical cartoon scrubbing bubbles." But then again, this is marketing we're talking about, which leads to a strong movie recommendation: How to get Ahead in Advertising.

More than you ever wanted to know about soap after the jump!

Wednesday, November 4, 2009

A stretch

I am now full and well wholly exhausted. Today saw Release 2 for software engineering; yesterday saw the most recent time I got out of bed. Far from complaint, there's something about staying awake for 36 hours or more that strongly appeals to me. I enjoy the feeling near the end of it, a sort of lightness of being. I enjoy the solitude of the earliest morning/latest night, that short period in which this little city is blanketed by an expansive silence--and Oh, those precious moments you can hear the sound of snow falling, a performance so minuscule that only amongst the stillest movements can an audience find itself. Further, with some associative tendrils linking them all, there is special joy in getting into bed after having not gotten into bed for some atypical stretch. It is almost as if as time goes on, the general average distribution contracts to a focal point of particular lucidity, a larger than normal indefinite fog lying within the perimeter that typically defines the periphery, and which invokes some new order of perception in many ways desirable but at least for its provocation of unique insight.

I Really enjoy the amazing amount I can accomplish over that time... I have some sense stronger than naught that I'm not really able to hit my stride until 12 or more hours after awakening. Actually, I recall clearly that despite finally managing to awaken early enough for class yesterday, I was in a Franciscan class mental fog nearly all day--it wasn't until around 11 PM that the urgency of the upcoming deadline and hopeless mounds of work (even worse, the mounds hadn't yet been assembled, by then there was merely a postulation that some mother-lode waited for a few rocks to be overturned) translated into some motivation to begin working on it. To be certain the thought of a good chance to pace my new furniture played some part, but before long, with some pride of performance and an awkwardly intangible form of irony, the chair disappeared and I became absolutely consumed with work. I worked for 17 hours, from 11 PM to 5:30 PM, and the only thing that stopped me then was the necessity of attending class; despite the fact, I was 15 minutes late for my inability to find a timely conclusion. I wish I could explain exactly why this was the case, but I'm afraid it's the type of situation that even another knowledgeable programmer would have a difficult time understanding. In sufficient, through the final pressing hour I managed to do what I would have previously considered impossible from a number of perspectives. That is to say I experienced some minor miracle nonetheless greater than maintaining consciousness through somnolence by programming a computer to move monkeys and boxes about a screen with bounded futility.

Actually, it was a lot of fun. Throughout the hours I happily dabbled slightly deeper into otherwise foreign but interest arts: graphical design, audio production, not to mention the manufacture of code that has a functional, visual interface.

Tuesday, November 3, 2009

Herman Miller Embody - first reactions

I think talking about "my chair" is banal and way too close to boastful, the egotism inherent therein being a personality characteristic I try particularly hard to avoid (synonyms of boastful read like a list of things I doubt many people aim for as a characterization: arrogant, conceited, pompous, pretentious, etc). With that in mind, I think "the chair" in general is fascinating, and am grateful that I have the opportunity to experience it in real life; though I was hesitant to write this up, perhaps my thoughts will lend insight to someone else.

With all that said, I have to admit, few things will sound more pretentious than what I'm about to say anyway.

First, the Embody has an arresting aesthetic, its form is absolutely captivating; it is actually inspiring to look at. I don't mean inspiring in the generic, feel-good way, I mean when I look at it there is a surge of creative, unique thoughts and a sense that the majority of objects we encounter each day are needlessly bland, expressionless, devoid of notability--altogether invisible and uninspiring. You might look at a picture of the Embody and think 'I don't get it, looks like a moderately interesting chair,' and with that I'd agree, a picture of the Embody displays a moderately interesting chair. However, as I've mentioned before, the 2 dimensional projection of a 3 dimensional image loses and incredible amount of information; the chair IRL is a whole different story.

Of course, function is a critical element of design... personally, I think form is just the whipped cream topping of any design. To be sure, a desk isn't much of a desk if you can't use it as a working surface, no matter how beautiful it is--Michelangelo's David is not a desk. On the other hand, as long as you can use it is a working surface, no matter how nauseatingly ugly it is, it'll work as a desk. Thus, it is a good thing that the Embody has function covered. But there are levels of functionality, and true to the reason I chose this particular chair, it seems to have function covered to an exceptional degree; not only can you sit in it, sitting in it is a pleasure. I haven't had the chance to sit in it for one of my 12-hour-straight coding jams yet, and that's the true test, so the full extent of its functionality remains to be seen. As it is though, it's a pleasure to sit in, it feels something like sitting on a bed. This isn't altogether surprising, as the seat has a system of suspension much like a mattress. This is a wonderful idea, and one that really surprises me for its obviousness, yet lack of presence in every other office chair I know of. The back also has an interesting suspension system, one with less give.

One thing that I love most about the Embody is that despite looking and sounding very complicated, despite a long design process with many prototypes, it is actually surprisingly simple, especially the suspension systems. The composition, strength, elasticity, and formation of the various plastics used is probably fairly involved, but in the end, the shape and intersection of all of them is very natural and efficient. Certainly this was a design objective (easier said than done), but the result is powerful; the Embody looks like an exoskeleton, an extension of the body, and what could be better than a solution provided by nature?

I don't think it's the be-all, end-all of office furniture, but then again I have a strong bias against conclusive permanence, so no chair will ever fulfill that criteria (except possibly an infinitely adjustable "indefinite chair" formed in real time by nanobots). In case you couldn't tell, so far I love it. The only shortcoming I've thought of, an insignificant and unimportant one, is that it doesn't have a headrest. There's no obvious reason why they might choose to omit such a thing, but I'm certain that it wasn't a simple fact of oversight--clearly there was no oversight involved in the design of this chair. The only reason I can think of is that they figured the inclusion of a headrest would provoke a change in the implied posture which would possibly have a detrimental impact on ergonomic functionality. Well, I'm sure there is a reason, I'd like to know what it is. My only other complaint is that it took a long 8 weeks to get here.

Sunday, November 1, 2009

A Pictorial Vindication

I have received word that, given the existence of infinite permutations of parallel universes, some people have been considering my recent furniture upgrade as unnecessary. However, in light of what I'm about to show you, I believe any transdimensional rumors will be sufficiently concluded as having no factual foundation. I present my current chair (with bonus Pickles action):

If one happened to feel gifted with a finely tuned, elite aesthetic sense, and that by seeing this image this sense has been dismantled and disfigured, transformed irreperably into an unidentifiable blop of goo, allow me to comfort them; had their sense been incapable of handling this fundamentally evocative display of Form, it wasn't worth a tarnished penny anyway and nothing of value was lost. I suppose they wouldn't see the profound symbolism in the bits of open cell gray foam that with clinging tenacity remain on the rough plane of misshapen splinter spitting plywood, refugees of a forceful division of useless padding from malformed foundation. Nor would they see, I imagine, the ongoing dialog between the thoughtlessly intrusive, precariously balanced seat, mismatched and standing as a contradiction to the firm stability of the back, which with a posture slightly less than vertical seems to mourn the loss of its intended counterpart. Of course it all goes so much deeper, more than I could fit into any number of blog posts.

I will surrender, yes, that I assembled this marvel in a fit of vanity, that I ought to have focused from the start on a more balanced, less perfect design. However, I assure you that the striking, fluid beauty of this piece is complemented by an enhanced functionality; namely, this artifactual collage doesn't make the lower half of my body go numb as its unglorified predecessor did. Frankly, I do admit some amount of guilt for purchasing a replacement for a chair that is otherwise the paragon of design, but I think that it is better served encased in glass as an enduring testament of what can be accomplished with concerted effort and a healthy dose of luck.

In case it sounds as though I've gone completely bonkers, the truth is yes, I have--but only temporarily. Such is the result of staying up all night doing probability homework.