This is a terrible idea. I can look at the code edits you did to xmrig and then see how they can easily be modified for xmrig amd. So what is going to happen is that those with the technical know how are going to do the modifications, use GPU miners to start mining at a higher hashrate and then those that can't will be left behind.
If this algorithm is not GPU resistant, which it isn't, you need to either make it GPU resistant or release a GPU miner.
what you said is simply not true.
https://cryptonote.org/cns/cns008.txtlet me quote
CryptoNight is a memory-hard hash function. It is designed to be
inefficiently computable on GPU, FPGA and ASIC architectures. The
CryptoNight algorithm's first step is initializing large scratchpad
with pseudo-random data. The next step is numerous read/write
operations at pseudo-random addresses contained in the scratchpad.
The final step is hashing the entire scratchpad to produce the
resulting value.
Also less technical but more practical explanation, with comparing power costs/etc.
https://monero.stackexchange.com/questions/1386/is-it-better-to-mine-with-a-cpu-or-gpualso since this thread was written on stackexchange, GPU prices went up a lot (especially those graphics cards that are good for mining), so I wouldn't be suprised if CPU mining, at least short term was actually more profitable than GPU if you include costs of acquiring hardware.
We never claimed that our coin cannot ever in any way be GPU mined, but
1) It won't give GPU miner big advantage, it will be quite small
2) we will not support such miner
if someone creates it, great, but we will keep changing algorithm from time to time and he will have to keep rewriting it.
and all this for not so big profit compared to CPU mining? It still might happen, but effects will be quite minimal, thanks to this algorithm.
Our goal, after releasing GUI wallet and many other things is actually to rewrite this algorithm to minimize GPU minining advantage to almost zero. At this stage, we will gladly release GPU miner ourselves