So, while you may have a world record at wasting money to look for prestige, I can assure you that you are definitely not a good coder (in my eyes at least, because I am not seeking anything from you, not even your out-of-this-earth CUDA skills).
@kTimesG I respect you a lot because you knowledge on Kangaroo, if what you said it is true this mean that with a real optimized version of kangaroo will need less GPUs to solve next puzzle?
So you really don't see anything fishy with these? There are only 2 options in my view, since RetiredCoder confirmed (roughly) my calculations on cost:
1. He has too much money but there's no actual "crazy fast" kangaroo (faster than double the fastest speeds we shared here). since no one can reproduce his claims in accordance with the implementation limitations. Hence, making fun of us.
2. Keys were known and its all a hoax with some loose cover up code and vague claims of some "K" in absence of any explanations, proofs, even tentative thoughts, etc.
Problem is on dealing with cycles on a GPU, oh well, let me just say that having even a couple of registers used for something else than performing some parts of some jump of some kangaroo, can decrease the speed by well over 100%. Not to mention the extra required access to memory when the registers are all full (to "deal with the cycles" one way or the other) bringing the speed down even further.
You just don't have enough skills in GPU coding.
Exactly. I'm a retired coder, a good one, so I do it for fun mostly when I have time and interest, and it becomes boring after a month of managing a network of GPUs.
I have a question, if this is just for fun, why do you keep the knowledge to yourself? I understand that we mortals keep things to ourselves to have an advantage, but how does that benefit you? And you will say that you shared kang1 but your "SOTA" code I have tested and implemented on GPU and it is even slower because it makes excessive use of ECC operations, while the ideal is to work with Int and only use points where strictly necessary.
so even though you think you are a good programmer you prove the opposite, which makes you even more suspicious of being the creator and you went crazy, staining your own lore.
When gpu test version?
It's part #3, the last one, so not very soon.