Topic: [GPUC] GPU Coin | Mandatory Wallet Update - page 249. (Read 421421 times)

even if there will be new asic that can mine new scrypt coin algo(keccak, n-factor ecc...) there will be also new algo, which asic will not be able to mine, in the end there will be always at least one algo that is anti-asic...deal with it asic are bad

I wouldn't put i past ASIC makers to be able to get effectively into this n factor increase market, like i said it's all about how money there is in it...especially for adapative nfactor, as it is really is just a small evolution.  Fast memory controllers isn't a walk in the park though.. some GPU businesses have lost millions due to it.  It's tricky stuff, not everyone will have the mojo to do it.. even if they have a 25nm chip fab in their home..

Algos are evolving, adaptive-nfactor is a small evolution on that front.  I would not put it past the abilities of the community , if we really wanted to.. design an algo that is very specific to the GPUs.  Maybe run your hash through some 3d tesallation of the model of that monster you just illed in quake.. and then back through an h264 encoder... or... something...

To network be allive we don't need HIGH speed in TGS speeds... just because we don't encrypt harder this is no point in making it for ASICs. Asics are for people who want exploit crypto and they will destroy idea of decentralization...
GPUs was chosen just because all PC users have GPUs...
And cpus are using botnets so Gpus left as right choice.
GPU coin farmed by ASICs this is REAL JOKE guys.

its doable, just because no one sat down and focused on making a professional grade home chip fab yet doesnt mean it will take years to do so. diy chip fab actually already exists, ofcourse its nowhere near capable of 65nm or let alone 25nm.

if gpu makers do get into the game, there wont really be any difference to the end result of hashrate monopoly's. your worse off with gpu makers getting into crypto mining then having asic companys get involved.

If it's not - n, at the moment this coin are extracting by intelligent mine pool  from China. Only the value of this coin are
 destroied.

It doesn't matter if the algorithm is Scrypt or N-Scrypt, it's the same conditions for everyone. The block timer is still 60 seconds and the coin reward is still 20,000.
Let's say you have a GPU rig that amounts to 10% of the total hash rate for the coin. Regardless of algorithm, you would generate, on average, 2,000 coins every 60 seconds.

We shall see about that.  maybe some day i can start printing 25nm chips in my home... but then by that time we'd probably be using quantum tech..


I think you're assuming that the 2 GPU companies aren't watching or don't care about this market...

this is going to be proven wrong real soon, professional diy chip fab is coming
once i get my manufacturing capabilities online, its actually a project i plan on tackling.
also gpu are stuck to their gaming purpose, an asic can be designed to be adaptive proof with more memory then you would ever find on a gpu, heck you could build an asic that can also interface with pcie and make a hybrid. if your coin doesn't make it to the asic point, its never gona be worth much.

N = 2048, i think ASICs have a shot.. but as N increases.. I think ASICs will have a hard time staying on top of GPUs, as long as there is still a market for GPUs (and mobile SOCs don't overtake everything...).  GPU development is state of the art right now: Memory controllers, FB sizes, latest nM processes.. the engineering capabilities at a lot of these startup ASIC companies in alibaba will be l.o.l.  

Maybe one day the coin asic market will be bigger than the GPU market.. and hence they will have the engineering brains, process-capabilities (25nm go!) , wafers at TSMC.. that GPU companies have a monopoly over in semi-conductor industry.  

But given that today, the estimated market for this is ASIC coin market is $100M... and the estimated market for GPUs is at least $10B... i just doubt they will be able to keep up if algo difficulty rises.  

But .. hey who knows..

Received my IPO email a couple hours ago and must say that you have alleviated some of my frustration if not all. As I read I really liked what you were saying and completely bought me back into the coin

-n isn't going to make this coin exclusive to gpu, its just a short delay. they are already in production. you can contact any asic manufacture from alibaba.com and tell them "i want a miner designed to mine -n exclusively" throw them money and bam, you've got your -n. just takes one person with a paypal account.
timeframe on -n miners can simply be bumped up by throwing in clauses for deadlines and the sooner they get it to you the more reward the designer/manufacture will get. throw enough cash at the problem and you'll have your device by the end of the month. this is personal experience from using alibaba.

since the devs are fighting so hard for -n i'll gladly take back my ipo. no sense in debating this topic anymore.
you've officially killed this coin, congratulations!

truth is asics can do anything they are programmed to do. out of all the "resistant" algo's I do prefer keccak and so does the lifetime of my gpu's :/

Ultimately if there's a market for it.. ASICs will be superior in whatever app they intend to service... That's what their names imply... Application Specific .. integrated circuits....

However, the idea here is to make it not worth their while to design, develop and fabricate the ASIC.. asic development isn't like writing software where you get 13 guys, and sell a company to FB for a billion.. it takes real capital to fund R&D, buy wafers and fabricate.. lot of risk.. risk wafers, imperfections.. then debug, ECOs and respins... it's a big upfront capital investment.

The nfactor code isn't just a set algorithm.. well it is, and it isn't... it changes 1 parameter on a scheduled basis by posix time to be "adaptive"... okay its pretty lame adaption change just 1 parameter... but hey it does increase the memory b/w dramatically.. 1024 -> 2048.. and on power of 2s... which *might* make it infeasible for ASICs to continually tape out to adjust for it..  and if ASICs should choose to also become adaptive.. well.. they just become generalized processors anyhow, which would yield little to no advantage against GPUs, since they are no longer really application specific... if you support 100+ applications.. well then you're a application processor..

Sure... if at some point nfactor=X has a 100B market cap due to coins.. and its just worth people's time and energy to keep developing and releasing asics that do just that... then yeah, this premise does fall apart..  But the lead time and development time is long for integrated circuits.  Idea here is to not make it worth their while... It may fail.. but in my estimation it ultimately it all depends on where crypto market cap in total are going...


Thinking about this.. there's probably room for a more innovative solution to this which not only changes nfactor.. but changes the algorithm based on blockHeight... to align with the direction of where GPU/CPU computing is going.. like for blockHeight > 10000 or something.. swtiches to prime sieving.. (primecoin)... predicting in that time GPUs would catch CPUs in that race in a few years.. and leave ASICs far behind... but anyways who knows, the future is the future.. anything like this is hard to predict.  If you could you would be making millions speculating the semiconductor market

I am aware of the differences in profitability for mining. I am talking price per coin (this is more for IPO investors)

Dogecoin 100 billion coins. Price per coin = ~ 160 Satoshis
Litecoin     84 million coins. Price per coin = ~ 0.024BTC

Finally an official update... Project is not a scam and the dev decided to go further with micryon and cryptowest. Wish them good luck and let's hope for finally a successful launch of this project!

you not aware that doge is actually more profitable to mine than litecoin?
that one coin alone has proven that amount of coins doesn't matter, its the majority that makes it valuable and using -n will tank this coin.
if any changes are to be made, it should be the IPO holders that have the say, not some stupid 2hr poll on a forum with a 10000000 to 1 troll ratio.
I've got personal investment, why should some random person get to vote when they likely aren't even going to mine it?
the general consensus with the IPO holders was "we dont want -n"

This is all true.. however.. you have to keep in mind that this *is* GPUcoin.. for people who want to use GPUs..

ASICs will be released for scrypt.. and although they won't be huge orders of magnitude better vs. what we saw with bitcoins, I have no doubt they will, over time be better at perf/watt and price vs GPUs for mining scrypt.  Staying with scrypt will mean that GPUs will eventually become obsolete in this race for mining , as long as this whole thing keeps going to develop it.... i gaurantee it..

Nfactor schedule: while it is absolutely true wastes more electricity, and silicon area.. but it keeps this a GPU mined coin.  The overarching assumption here is that we that follow this thread and mine it , *want* to use GPUs to mine this coin..


Now that point is a larger debate, which I think is really outside the context of GPUcoin itself... and i'm sure we can rathole on that topic endlessly...

Listen, people. NOTHING is ASIC resiatant in the long term. When LTC was created, using a scrypt algo, it was meant to be ASIC resistant. However, technologies go further with high speeds and 2-3 years later we do have ASICs for scrypt.
Understand that whatever algo to be used, soon or later the asics for it will come out. One may take longer, but soon or late, ASICs (or even new type of hardware) will be created for anything. Yes, these algos are a short term solution. However, in the long term this is an everlasting war with two parties.

Supply and demand. If this coin is harder for everyone to mine than it will cause the price to be higher as a result. Look at alt coins that have a cap in billions. Then look at coins that have a cap of millions. What do you notice?

It CAN.