im a registered nurse don't know shit about coding. but i hate fuckin banks. and my research and passion for crypto privacy led me here. im glad to be a part
@0xb100d unfortunately grin wont get the same opportunity as bitcoin to organically grow so it must be carefully constructed.
@Mathias_Nguyen_twitter grin has an opportunity to grow more organically than bitcoin
@ignopeverell@tromp not sure if this is the best place to have a conversation with you guys, but I think it'll be worth having a longer discussion about your recent PoW changes
I could be misunderstanding your goals, but I don't think the current PoW proposal aligns with them very well
Both cuckoo32 and equigrin are accessible to ASICs, and equihash has enough implementations out there already, some of them designed to be flexible, that I would almost suggest that equigrin is more likely to have day one ASICs than cuckoo32 is
Especially because Innosilicon has already stated that they can update their current ASICs to different equihash parameters with a simple firmware update
I haven't looked to deeply into the equigrin parameters, but on their face it seems like you could create a completely memoryless ASIC for equigrin, and just eat the brute-force penalty while still coming out >100x faster than GPUs
And as it's blake2b based, which multiple manufacturers (Obelisk, Bitmain, Innosilicon) already have hashing cores for, the step between that and a memoryless ASIC for equigrin is relatively small
I had originally planned on pointing this out on the forum, and maybe I still will, but I think it'd be good to have a real time conversation, a lot of the things I'm reading from you guys don't match my understanding of the reality of the mining world
For example, choosing to exclude 6GB GPUs from the GPUs can mine means you are basically eliminating all gamers
It also means that you are narrowing it down to just a few models of GPUs, which are going to be more accessible to large purchasers with existing relationships to Nvidia than any smaller mining farm, your choices basically heavily favor the existing entrenched mining farms.
Well, that's assuming you don't get day 1 ASICs
But even if you do get day 1 ASICs, you aren't likely to find out: you've more or less publicly stated that you are going to be doing changes every 6 months to deter secret mining. If I were a secret miner, I would definitely take greater steps to hide my activities and go through greater lengths to make sure the ASICs I build are flexbile
Though it really wouldn't matter... 6 months is long enough to pay for a tape-out and subsequent manufacturing run
@tromp I specifically want to call you out on your dreams of cuckoo32 ASICs that are composed of some specialized computational chips next to memory
that's not going to be the fastest way to build a cuckoo cycle ASIC
going off-die is really expensive in terms of both speed and electricity. You get the best performance when you've optimized your die to spend as little time as possible off-chip, that means that every chip you build is going to have as much memory as possible, and also is going to have the computational cores right there next to the memory
At 28nm, a cuckoo32 chip network is going to be 2 chips positioned right next to eachother (potentially on top of eachother) that are basically identical. At 7nm, a cuckoo32 chip is going to be a single chip
Even at cuckoo35, it makes more sense to have a bunch of identical part-memory part-computation chips rather than a computation chip next to a bed of off-die memory
@DavidVorick_gitlab are you working an an asic with grin in mind already?