so ..,asic still can be produced to mine this coin?
The algorithm takes quite a lot of memory, 256kb per thread, so its from the ASIC perspective closer to Litecoin/Scrypt than to Bitcoin.
The Scrypt miners took years to get where they are now, rough estimate: factor 100 (same power consumption). So thats 3.5TB with 200W of power in BURST.
i wonder whether the passphrase ppl are using has any baring on the luck ppl are experiencing. i for one used lastpass to generate a 100 character passphrase consisting of only characters that look like random combination of words joined together.
I can assure you: The passphrase has nothing to do with your chances of finding a block.
About the GPU issue:
I've just finished an OpenCL-Test-Implementation. On an R280X it takes ~22s for 10k nonces, assuming the final hashes are pre-calculated. With a block time of 5 minutes that GPU equals ~35GB of disk space while consuming 200W of power. No real alternative.
http://www.jocl.org/
This library offers Java-Bindings for OpenCL that are very similar to the original OpenCL API. The functions are provided as static methods, and semantics and signatures of these methods have been kept consistent with the original library functions, except for the language-specific limitations of Java. The OpenCL API may be very verbose at some points, and this is not hidden or simplified, but simply offered by JOCL as-it-is.
Why even use Java? It'll be a small C tool that generates exactly the same files. But - its still far from being done.
