Beginner Distributed Process Project
For the longest time, I have been interested in creating a cluster of heterogeneous nodes while trying to build a home supercomputer as I am very interested in AI research.
However, the problem is that I have a lot of hardware (2x rack mount dual core servers, 8 285GTX Gpus, 6x PS3s 2x Hacked 360s (they can run linux) access to tons of regular PCs as well as multiple workstations) there is no big dataset to wrinkle, or even any software that I can run. I messed up with a distributed code compiler, but at best it made my kernel builds from 10 minutes (worst case) to 30 seconds (and I think 20 of those seconds are just set up).
So where do you start? I have a decent understatement of Obj-C / C / C ++, so it's not worth writing too much, but what should I write?
a source to share
If you wanted the data to crunch, there were many:
- Data Range and Knowledge Discovery Data Sets
- multiple scraper and / or replaceable datasets
- Comprehensive Knowledge Archive Data Pack List
- collection of big health data sets
As for "what should I build", the real question is, what are you interested in?
a source to share
Well, I think it's best to determine what subset of hardware you have that you will be developing your application for. PS3 software requires special attention and will require separate development from something built to run on typical Linux servers.
You may also need some research on how you can develop a 360 app; I'm not sure if this will really give you what you are looking for, to be honest.
Once you've decided on a subset of the hardware you need to develop, it might be a good idea to start with some basic development so you can build a foundation that allows you to communicate. With a solid foundation, you can extend your code to support a wide variety of distributed projects.
Hope I understand your question correctly!
Greetings
a source to share