Odroid C2 cluster - What to do with it?

Great looking system. I can't make a knowledgable suggestion, but I am interested to see what you end up trying.

Alex Ellis had some good posts on running Kubernetes (or recently k3s) on a cluster of Raspberry Pis. That would seem like an obvious start.

https://blog.alexellis.io/test-drive-k3s-on-raspberry-pi/

As for what to run on it... it seems more like a learning exercise then anything practical.

Do you have any more info on your ECM factorisation? That sounds interesting.

Fred27  wrote:

 

...

As for what to run on it... it seems more like a learning exercise then anything practical.

 

Do you have any more info on your ECM factorisation? That sounds interesting.

I agree with Dave (or FRED?) it is a great learning. I too am curious of what the ECM yielded and time devoted.

From Wikipedia: "The Lenstra elliptic-curve factorization or the elliptic-curve factorization method (ECM) is a fast, sub-exponential running time, algorithm for integer factorization, which employs elliptic curves. For general-purpose factoring, ECM is the third-fastest known factoring method. The second-fastest is the multiple polynomial quadratic sieve, and the fastest is the general number field sieve. The Lenstra elliptic-curve factorization is named after Hendrik Lenstra."

I use the Loria implementation of ECMNET, a project that distributes the search along many different computers. As the name states, the algorithm runs "tests" over elliptic curves to find a particolar configuration, whose meaning resembles the work of trial-factoring. Each clien runs GMP-ECM a program that executes the factorization tests. GMP-ECM has versions for INTEL/AMD processors, ARM archtiectures and CUDA parallel systems.

The Elliptic Curve Method is seen as a generalization of the P-1 algorithm, an extension of the Pollard Rho method (here is an implementation for CUDA systems).

I have been interested in computational number theory since my purchase of a Commodore C64 (my first program was the calculation of Armstrong numbers: the progam written in BASIC took 3 days to complete).

Now I explain algorithms, implementations and other projects related to SBC on my blog. Here is the Math archive.

Luigi

---

Run Seti@home... Maybe find alien signals.

I'm a simple Minion, I'll try to
1. Compare its performance with lowest power intel/amd cpu.
2. Make a fart gun out of it, if possible
3. Give it a job, Cryptography or cracking a password.
4. Use it for Machine learning/AI. make a Bot out of it.

cheers!!

I didn't know they had an ARM version of their software!

I already completed section 1, and published my results on my blog.

Cryptography would work better on Jetson Nano, with his 128 Nvidia parallel cores. ANd yes, I am buiding a small 3-Nanos cluster to do it.

The same works for ML, as the Jetson Nano already has parallel PyTorch systems installed.

I was looking for something "different", more distributed and less number-crunching intense.

Can you make a Cookie robot out o it?
Ummm... Home security or automation? with multiple cameras?

Home security is my second choice.

I was thinking to create a (somewhat dated) dashboard control of the old NGC1701, the Enterprise