It is very hard to predict the next state of computer science. Especially, after ten years. But, i think there are some clues i can play with and i am going to make some assumptions about the data mining solutions.
Currently in the company i am working for, we are trying to create some intelligent interfaces for our business customers to make them play with some huge databases as fast as possible. But obviously, it is getting harder every day. The amount of data we are working on is growing too fast. Faster than the solutions we prepared or the computer technology the market has.
Certainly, the problem i mentioned above is not just a problem for my company. Google to Microsoft, all the companies have some sort of problems around the data mining. At first, I can say that the position I am working on will be important as much as now in the future I am going to predict.
Actually, to make better assumptions about the next 10 years, i will compare the present with the past 10 years. I remember the server pc of my former company. There were 3 SCSI Hard disks installed inside the machine and each had 18 GB storage space. That was a good amount of storage capacity to work with. Even today, I know some guys working for that company and I am sure that the storage capacity they need, 20 times bigger than the past. I am not going to talk about the reasons for the capacity. But the obvious is that they will need much more storage tomorrow.
It is not just about the storage capacity. It is about the performance we demand. RPM rates of new hard disk devices or calculation speed of brand new CPU cores. Every part is important. According to Intel’s SVP, they can predict their next 10 years. But, i am curious about the explanations. Aren't they very anxious about their stock prices?
As David Liddle said, in 70’s software researchers found a way to work with parallel machines. It makes me think that, maybe there are some limitations with the devices made by current technology and maybe we reach to end of it soon. But, on the other hand, software developers will find out some faster algorithms or techniques to achieve the needs in more efficient way.
As a data miner, I can say that we will be needed to make some operations faster soon. We will be needed to find another ways to achieve the future goals efficiently. Probably we will be using faster and smaller computers than now. But not that fast to handle the data available tomorrow.
References:
The Wider Impact of Moore's Law
http://www.ieee.org/portal/site/sscs/menuitem.f07ee9e3b2a01d06bb9305765bac26c8/index.jsp?&pName=sscs_level1_article&TheCat=2165&path=sscs/06Sept&file=Liddle.xml
Does Moore's Law Still Hold Up?
http://www.edavision.com/200111/feature.pdf
Moore's Law: "We See No End in Sight," Says Intel's Pat Gelsinger
http://java.sys-con.com/node/557154
As an answer for DQ at University of Liverpool - October 2010
0 comments:
Post a Comment