I don’t like politics. I don’t enjoy taking sides(in the political realm). However today I was taking a quick glance at Hacker News and guess which political figure’s name popped up on the site I go to for TECHNOLOGY news. If you guessed Obama, then you, my friend, are right. Now, before I continue, read the post by the BBC which gives you an overview of what he wants.
Supercomputers: Obama orders world’s fastest computer
If you haven’t realized by now, the United States of America’s presidents (all sides) know how to spend money, none of them know how to make it. You might also look at the American deficit, and try to figure all that crazy stuff out. Yet, one thing the USA does know how to do well is compete. Remember the race to the moon? Yeah, we won that.
Yet now we are on another journey, a race in the technological age. However, and most unfortunately, we are still confined to the one track mind of our leaders. Obama isn’t the most technical person in the world, and we know it. That’s Okay. You don’t need to have a degree in electrical engineering and computer science in order to run a country. What might be helpful, however, is finding good E.E. and C.S. advisory when it comes to a technical decision. The ability to consult them upon a large costly project could help save a ton of money in the end. If it doesn’t, then maybe your advisories aren’t as good as you thought.
If you haven’t read anything about China’s latest super computer, do so now. It can compute around 33.86 petaflop/s and they call it the Tainhe-2. Oh, and then there’s the cost of the energy it uses: ~$400,000 USD per day.
So what is wrong with this massive behemoth of a machine? It’s simple, no one uses it. Most super computers are built to do one thing very well; this is abnormal when compared to your everyday computer. Your home PC was designed to do a lot, well enough. It processes graphics, does arithmetic with integers and floating point numbers, along with the ability to save data to the hard drive for use at a later date. Armed with this knowledge, developers target their software towards everyday consumers, and anyone who has worked with a software development company knows the cost involved (hint: it’s EXPENSIVE). Therefore software engineers focus on consumer equipment in order to maximize their programs’ usability. To get a program working on a super computer, one must pay highly specialized developers in order to create a program that might only be used once. It also might take a long time to build considering the missing tools which are available to use when targeting the consumer side of the market.
Let’s stop and think for a moment. I know pride causes rash decisions that can ultimately result in some happiness, but it can also inquire cost. I propose that we stop this ill-thought idea of being overly prideful; I get it we are American’s. I love being first. Who doesn’t? The fact is, we need to be more strategic. We need to show other countries that we are capable of thinking outside of the box. And just like science has already shown us… we can.
There are programs that allow you to dedicate your computer, in its spare time, to research teams to do a few computations in which it processes and then spits back out and pushes that data to a researcher’s server. Although servers can get expensive, they don’t come close to the cost of renting a super computer for a day. Quite a few Scientist even host their own. In fact, here are a few that you can contribute to yourself. Most of them are for a good cause.
Now take that concept and apply it to a well-rounded super computer. We could have a centralized server which serves up data to volunteers computers and retrieves it upon computation completion. Think about how much money that would save (specifically tax dollars). So rather than wasting efforts on a super computer to please our president and our Egos, why not invest in a smart, distributed system of software that would save money and be just as powerful if not more so. The software could utilize a user’s spare GPU or CPU time and send the results back to its headquarters. It is a wonderful idea and reality that can happen and hopefully will (in time).
With that stated, I ask these two questions: “Why are we wasting our time and money?” and also “How can we use innovation to waste less time and money?” Feel free to respond below in the comment section.
Shoutout to @davidvkimball for his help editing this post