CPU friendly Altcoin in development

[Lauda]

Still waiting for this 

For what? CPU friendly coin already here, it called prime dice.

Whaat? Link.

[minerapia]

dunno how heavily youre coin is based on radixsort but theres an example for it in opencl, so It's prolly not that hard to make GPU miner.
https://developer.nvidia.com/opencl#oclRadixSort

[msweb]

dunno how heavily youre coin is based on radixsort but theres an example for it in opencl, so It's prolly not that hard to make GPU miner.
https://developer.nvidia.com/opencl#oclRadixSort

It's far away from being a coin. Thanks for the link. It seems I have to take a deeper look at current radix sort calculations for GPUs. Reading about it seems to me that radix sort isn't the way to go here in the long run. Scrypt with fixed parameters isn't very wise too in the long run.

[FreeTrade]

The model I favor is rapid distribution followed by a 2% inflation. I'll have more about this soon.

Okay - so in case you hadn't guessed - this is now MemoryCoin. I think it is the first really GPU-Resistant Coin.

[barwizi]

you people truly do not understand, you can use new methods but all you do is delay a gpu implementation. There is no such thing as GPU resistance and ASIC resistance. If the market and money behind any coin was large enough, all that would be in dev right now.

[Seth Otterstad]

Obviously you cannot write an algorithm that cannot run on a GPU but runs on a CPU.  What we are concerned with is making the ratio as low as possible.

[Seth Otterstad]

What about a proof of work requiring hard drive storage? Or even a lot of memory - it's expensive on server farms and hard for botnets too.

The memory route is scrypt.  The hard drive route doesn't work because you can have a million computers access the same hard drive.

[jbrower]

All-

I apologize in advance for posting on an old thread.  The initial section of this thread is relevant to my question and I haven't found other forums with an answer.

For Litecoin, if memory size is increased in processing the Scrypt algorithm, for example to 384 kByte instead of 128 kByte (and assuming likewise an L2 cache size of 384 kByte), would the typical CPU architecture (x86, PPC, etc) be likely to see a corresponding increase in performance?

I realize that increasing BlockMix memory usage is not typically done on GPU architectures, for various reasons.

Thanks.

-Jeff

[opticalcarrier]

i know its of topic, but can you tell me what im doing wrong in my altcoin creation?  I generated a merkle and also a genesis, but I have 2 peers configured with each others IPs as per instructions but they just dont see each other.  I tried it 2x, i dont think I missed a step. was there something missing from instructions?  I picked some random base58 address, the instructions werent very clear at all, but seems like that bas58 address really wasnt that big of a deal.

no im not going to release this here, its for another forums site I frequent, just want to give everyone there something to mine as a fun thing / joke sort of thing.