Parallel Computing on any Desktop

     So I was asked to read the paper Parallel Computing on any Desktop, written by Ami Marowka, which you can read here. The topic that Ami Marowka touches in this paper is really simple: It's all about accessibility. I'm a huge fan of the open source world. I love openness in stuff. I also like to get things that are super useful for free and I'm also a big fan of feeling that I can somehow contribute in others having things that work for them for a really low price (or non at all) too. And specially, that last thing is the one point that makes me feel a lot into the world of open source: the feeling that I can actually do things. 

     I think that's the point that Ami touched in this paper the most, the actual availability of such an important technology is no longer confined just in the headquarters of some super millionaire enterprise or in some university laboratories... the thing with this technology is that this technology is right there, for everyone to use. The thing is that we somehow assumed that things would naturally come (as discussed in the free lunch is over, that's no longer a choice). We have had multi core processors, with multi threading capabilities for a long time, we just didn't actually exploited or used their whole capabilities. Now we've come to a point in which we kind of have to learn how to do it. 

    I think that capability that we know all have to learn how to use this technology and extract every drop of juice out of it, with something as simple as time, is really important. Actually, it may sound silly but the problem with this new technologies is that, the time it takes for people to learn it, to cope with it. Parallel computing is not difficult at all, it just has a really steep learning curve. Is not like you could just pick up some random Python tutorial online and learn it in a Sunday afternoon. The state of the art right know is actually quite the opposite. Learning parallel computing, specially with languages that are actually made for parallel computing, will take away a huge piece of your time. A HUGE one. So the problem is that, the technology itself is really accessible actually, the problem is learning how to use it.

Diego Canizales Bollain Goytia