An ode to Sumpah Pemuda
The known fact on our country is that we are always at the consumer end. Take instance of the pharmacy industry. We are blessed with rain forest and genetic diversity, yet we are buying recipee from other country just to make Paracetamol. But, the pharmacy industry out there can make recipee by using software on supercomputers.
Why can’t we do it? Because the supercomputers are expensive and not just the price of the supercomputer, but also the operation cost. It was only visible to buy the end product than creating one.
Was? Yes, because we have grid computing and cloud computing. These two buzzwords are interchangeable at certain point and I think in this level of knowledge, the two are the same. Those provide high performance computing, enabling us to model problems like simulation and do it with high precision. Another instances that can be achieved by high performance computing (HPC) is to calculate and estimate climates, forex, genetics (DNA), simulation on physics (including modelling a good car), and so much more including the now trending semantics-based science.
Indonesia has one, at least, it is InGrid (INHERENT GRID). The InGrid that I know is consists of Hastinapura (UI) and NL Grid (UGM). INHERENT is a higher education network, now part of Jardiknas, that connects universities in Indonesia. Hastinapura itself is a cluster system which consists of 1 head node, 16 node (each node has 2 cores), and 1 portal node. I don’t know what of NL Grid, but I guess it top notch also.
InGrid has so many potential, yet it has a major flaw: No potential user ever used it extensively. I’m not saying that people that sometimes use it is not worth mentioning, they really do worth, but the people that extensively using it as a tool for their research material is lacking. Many that do have potential don’t even know it exists.
The InGrid system also suffer another problem: it is a limited system with limited resource. If everyone starts using it, the machine will never be able to cope. We need another resources.
Fortunately, FOSS has provide many tools to build HPC systems, such as OpenMosix, Folding@home, Globus, and so on. So many but not known very much to us. The most intriguing idea of providing HPC is by using Internet cloud itself. Take a look at Folding@home and electric sheep. We can participate on the project by donating our idle CPU, that is by allowing a program took some process background or running as screensaver.
How is it possible?
Because with today’s computing, we are only using few percents of our CPU and letting some CPU cycles idle. So, instead of wasting some cycles to spinlocks, why not giving it to compute some data, for example a DNA data, a stochastic model, or maybe a batik generator? God knows what possibilities can be built into reality with that HPC.
There is a caveat, though, this can’t be done right now because of the following reasons
1 Unity in Field
I don’t know about you, but we did have a lot of ego in each university, even in each faculty. I realize, when I was computer science student, I would put a high esteem on my faculty and sought lower on others. We also had a student anthem (fortunately, never be taught anymore) that mocking other faculties. Anyway, you got the idea of how fragmented our world is.
Now, to use the system, we must take away that silly pride and put focus to the problem domains. Today’s application of science consists of multidiscipline practice. In order for a pharmacy can simulate a new formula, we needed a computer scientist to put a computational model, medical practitioner to have the formula, and natural science to have the molecular design — that is the minimum. In order to put a design on robotics, we at least need engineering and software people.To compute forex, we also needed macro-economics, mathematics, computer engineering, especially in database.
I hope you got the idea. We need a lot of people on different fields working on a problem. With scarce resource, we are not anymore making our progress exclusive as our field, faculty, or university achievement. We must join together and put national interest on the front.
2 Unity in Connection
There is nothing more severe than the fact that Indonesia have a very innefficient way of handling connection line. How awful? Well, for instance we don’t have a blueprint of fiber optic (FO) cable installment. It is common that when a vendor planted a new line, the worker severe another vendor that already put their FO. We are crowded with antenna to pick cellphone. Even, each bank have their own private network of ATM.
What is the cost of this (un/under)regulated situation?
We have higher cost of communication. This is normal, each communication vendor maintains their own connection. No bigger picture of how Indonesia can be connected in a better way.
Well, the government must put regulation and nationalize the communication line. This can be achieved because of the UUD’45 guaranteed that. We have socialism as our basis. But, the problem lies on beaurocrats and politicians. If they care enough, we could be like this, the regional government built their own FO based on referendum.
Another solution is to have our own cell networks, or RT/RW networks and we put our own liberated connections. This can be achieved by using wajanbolic for connection and use cable whenever possible. This socialism work is called gotong royong and can be achieved by many people.
Of course, the optimal solution is to talk to the telco companies and take them with us by letting the Net Neutrality and stuff. They may provide their networks for research and, maybe, donate some of their resource for it. The rest of the story, you can figure it out.
3 Industry Consensus
I have a worry about the tendency of our industry that copying anything that comes. The worst part is they don’t have enough consent of why continued investment in research is necessary. This is because a research is costly and not all of our industry is capable to provide one.
But, if the industry can adapt new business process, just like software companies around the globe adapting Free/Open Source Software, they can survived and innovate in something that common. The real problem of today’s society is we don’t trust in collaboration. We afraid of our original story get caught and getting copied.
Well, that’s why we need to learn about FOSS. Not just because of its software, but also the fundamental ideas behind it. With thinking out of the jailbox, you can see how everything is reasonable.
HPC computing is getting an interesting trends outside of this country, because it is a cheaper way with high probability of getting successful than using real experiments. But, without any unity, it can never be done.
Salam Sumpah Pemuda.