As Prices Soar, Graphics Card Manufacturers Appeal To Cryptocurrency Miners

[problematiq]

It is different yes, but it only takes a financial incentive for a company to start kicking out ASIC boxes, like a jump in a crytocurrency's value. As far as ASIC resistant goes, I high doubt it will take much for someone to over come that minor hurdle, there are other cyrptocurrency's that were resistant to ASCI mining, (just look at ALT coin) but that ultimately did not last very long.

What makes you think it's a "minor hurdle"? Not saying it couldn't happen eventually, but I don't think it'll be the same as for Bitcoin. Also, Eth is eventually switching from a Proof of Work scheme to a Proof of Stake scheme, which I believe is supposed to make it less important to have oodles of hashing power, and therefore would seem to make ASICs (and ASIC development) less attractive. Unfortunately I don't really understand the PoW vs PoS think that well.

Not sure what you're talking about with altcoin, given that altcoin is not an actual currency but rather a blanket term for cryptocurrencies other than bitcoin.

Sorry "Just look at the alt coins" It really is not a big hurdle as most people think, they hear "resistant to ASCI miners" and what they think they hear is "almost impervious to ASCI miners". VERTcoin was one that came to mind that was "ASCI resistant" that in the end only took a few months before there were ASCI boxes flooding the market. Granted, it is a POW currency as is eth currently.

Stop saying ASCI !! ! !

Long freaking day yesterday. ASIC

 

[redgarl]

The debate is more interesting than the article. So basically, my crossfire setup would have been nice for mining.

Unless you pay the kWh below 0.07$, it is a worthless hassle. 2 1070 GTX would come to close to a 0.750kWh... not to mention SLI only work with 2 cards on Pascal. Basically, it means you need 4 setups for running 8 cards.

It's funny because Threadripper and EPYC were designed especially for multi-cards applications for deep learning and Neural Networks development. Nvidia and Intel shot themselves in the foot by not seeing it... and it can also be extrapolated to mining.

 

[Faux_Grey]

Mining with my gaming computers paid off my Ryzen build, but it's pretty pointless at the moment, in south africa you can't get any decent GPU for love or money.

The addition of these "mining" cards wont help at all, as people will go for traditional cards for their resale value. Unless the mining cards are in huge availability and half the price of a normal card, they won't be bought.

Then again, crossfire/SLI a mining card with a card capable of display output?
Cheap way to make supported games faster.

 

[wiyosaya]

the whole damn thing makes no sense. friend of mine recently got a rig up and running with 8 1070s and hes making 30$ a day. thats a <mod edit> of effort for what seems like a little return. and if it crashes youve got a lot sitting there doing nothing.

I think you may need to read an investment wiki or two...
...
That's over 100% annual return rate....If you don't understand how ridiculously good that is, you definitely need to read an investment tutorial or two.

It's a little more complex than that when you are discussing mining. What is making him $30 today could make him $3 tomorrow. If it crashes or suffers component failure, that impacts your profit, especially if you aren't around to babysit it. They also start ramp the difficulty once a coin gains popularity, cutting into your mining rate. When talking about Eth specifically, in a few months they may end PoW and implement PoS. The inevitable shift would render all his crunching power useless. I believe his has time to make his money back and then some, but after that, who knows. Certainly won't be quite as profitable as you are projecting.

+1 for the poster with a level head. No investment that looks that good will ever return that much.

My guess is that anyone who is willing to put off purchasing a GPU while prices are outrageous will benefit from the latest cryptocurrency fad.

 

[TJ Hooker]

Unless you pay the kWh below 0.07$, it is a worthless hassle. 2 1070 GTX would come to close to a 0.750kWh... not to mention SLI only work with 2 cards on Pascal. Basically, it means you need 4 setups for running 8 cards.

How'd you come up with the 0.75 kWh value?
Also, SLI is irrelevant to mining, you could definitely have more than 2 Pascal cards in a mining rig.

 

[Cztnights]

Agree with problematiq 100%. Nowadays if you want to mine any cryptocurrency asic miner is the only way to go. And even then it is hard to make any sort of profit, when dealing with bitcoin. My choice if I was to start mining atm would probably be Antminer S9 or S7

 

[TJ Hooker]

Agree with problematiq 100%. Nowadays if you want to mine any cryptocurrency asic miner is the only way to go. And even then it is hard to make any sort of profit, when dealing with bitcoin. My choice if I was to start mining atm would probably be Antminer S9 or S7

This thread has nothing to do with bitcoin mining, and eth mining ASICs don't exist...

 

[DerekA_C]

all i know is I missed this dam band wagon just like the last few coins as an AMD exlusive GPU owner I didn't buy my 580 in time now it is to freaking high of a cost but would be nice to have built a 7 gpu system just for this have all the specs and it would be easy to do imagine getting 7 580's at the $240 price to out perform the $350 $400 1070 a lot more return in profit there but dang it hate being the last one to hear about profiting, so now i must wait for the next coin this time I'm waiting and paying attention to at least make a few grand off it.

 

[10tacle]

When will the prices bottom out for the 1060? The 760 gtx card I purchased last year from ebay recently died and am using a 460 gtx 768 MB card that is showing its age.

grab a 970 or 980TI for less and same power for gaming.

Even prices of used 970s and 980s have increased on eBay. Last month they were going for about $150. Now I see them in live bids that are currently ongoing at $225+ and the bidding is not over. I'm seriously thinking of unloading my 970s now and just springing for a 1080Ti which appears to be out of the target zone of miners. If I can get $500 for both my 970s, that's an affordable $250 upgrade to a $750 1080Ti.

So those of you who were on the fence of upgrading to Pascal and eyeballing either a 1080Ti or 1080 upgrade, now is the time to offload. This mad rush on cryptogreed is not bad for everyone. Heck for kicks I checked what used GTX 680s were going for since I still have an old rig with them in SLI. well over $100!

 

[quilciri]

It's a little more complex than that when you are discussing mining. What is making him $30 today could make him $3 tomorrow. If it crashes or suffers component failure, that impacts your profit, especially if you aren't around to babysit it. They also start ramp the difficulty once a coin gains popularity, cutting into your mining rate. When talking about Eth specifically, in a few months they may end PoW and implement PoS. The inevitable shift would render all his crunching power useless. I believe his has time to make his money back and then some, but after that, who knows. Certainly won't be quite as profitable as you are projecting.

This paragraph can be boiled down to the sentence in my original post "assuming the value of etherium keeps pace with the difficulty of mining."

 

[Dugimodo]

I googled a profit calculator for ethereum and calculated it for 2 x 1070 cards at 0 power costs and it's still not worth doing. I think this ship has well and truly sailed and the rush on graphics cards will be short lived when the miners figure that out.

I mentioned lightcoin in passing previously but just to elaborate it was also designed to be mined on graphics cards and resist ASICs. That didn't last they just made new ASICs, it was good for a while but I missed that one too.

 

[quilciri]

I googled a profit calculator for ethereum and calculated it for 2 x 1070 cards at 0 power costs and it's still not worth doing. I think this ship has well and truly sailed and the rush on graphics cards will be short lived when the miners figure that out.

I mentioned lightcoin in passing previously but just to elaborate it was also designed to be mined on graphics cards and resist ASICs. That didn't last they just made new ASICs, it was good for a while but I missed that one too.

I don't know what you consider "worth doing". Using this calculator: https://www.cryptocompare.com/mining/calculator/eth?HashingPower=27&HashingUnit=MH%2Fs&PowerConsumption=140&CostPerkWh=0.01

A single un-optimized 27MH/s 1070 with free power will generate about $110/month in revenue at current value. Like I've already posted twice, assuming the value of etherium keeps pace with the difficulty of mining, you've made your investment back in 4 months at current used 1070 prices.

Sell your mined Etherium each month, and even if the coin goes the way of the dodo after only a month (extremely unlikely), you've broken even, as you now have $110 and a card worth $350.

 

[Dugimodo]

Seems I did something wrong, came out about $10 a week when I tried it. But assuming it does work out at $110 and does keep pace (highly unlikely based on other coins in the past) in my country a 1070 currently costs ~ $850 and power is not free but more like .25c a KWh. I still wouldn't consider that a good investment just based on my prior experiences with Bitcoin and lightcoin. Ethereum has been around too long IMHO to jump in now and expect margins to continue.

Historically as difficulty increases and more efficient hardware becomes available mining with graphics cards has become progressively less profitable. I don't expect this one to be any different. My lightcoin experiment happened at a similar point to what ethereum seems to be at now, I made good returns for a couple months running 5 graphics cards but after 6 months I was no longer covering power costs and sold all the hardware. I did not break even but I came close. Also when it comes time to sell your old hardware you are competing with everyone else trying to dump it for the same reason.

I know it's not the same exactly, but just out of interest for lightcoin I spent around $3500 on hardware, about another $1200 on power. I made around $2k from lightcoins, managed to get about $1.8k back for the hardware (the graphics cards were hard to sell and I didn't get what I wanted) for a net loss of $900 (I did keep some parts though so not quite as bad). I enjoyed the experiment but concluded it was not worth the time and effort. To convert to $US just assume around 2/3 of these figures.

It's entirely possible to make good money mining, but there is definitely risk involved and no guarantees. Go into it with your eyes open and don't spend money you can't afford to lose.

 

[calken]

I see allot of people talking about profitability and return on investment in the hardware.

The ROI to the purchaser is not in question, it's the ROI of the ASIC producer. They have to sell sufficient quantities to cover cost and GPU miners will change currency when they are no longer competitive due to an ASIC alternative. This results in a drop in value of the currency, which is subject to the currency strength, resulting in a potential sales reduction of ASICs.

 

[bit_user]

GPU miners will change currency when they are no longer competitive due to an ASIC alternative. This results in a drop in value of the currency

I'm skeptical how much the currency will drop from GPU miners fleeing. This assumes GPU miners are sitting on quite a bit, and they're all dumping it in anticipation (or reaction to) the others selling out. However, once that price dip is over, I'm guessing the price will tend to recover towards its previous trend (or possibly higher, if less is being mined).

 

[joseph1968]

I think AMD will not rush to manifacture more as they want the focus on vega.
I belive that after the "gaming Vega" is out and orders are as expected,we will see all rx series back to stores.
NVIDIA is enjoying to see its inentory sold out and need the time to focus on next gen..
Win win exclude end users.?

 

[bit_user]

I think AMD will not rush to manifacture more as they want the focus on vega.
I belive that after the "gaming Vega" is out and orders are as expected,we will see all rx series back to stores.

I think Vega is strictly high-end. It's intended to replace the Fury cards - not the RX series.

If I'm wrong, please educate us.

 

[Karlos1544]

GTX 1080 are no good for ethereum mining, which is why they're unaffected. Their memory's is incompatible

 

[quilciri]

GTX 1080 are no good for ethereum mining, which is why they're unaffected. Their memory's is incompatible

For clarity sake, the 1080 does mine etherium at around 21mh/s, which is less than the 27 mh/s of the 1070. This is attributted to problems with the GDDR5X in teh 1080 vs. the regular GDDR5 in the 1070.

For gamers with a 1070, this is a golden opportunity to upgrade to a 1080, as you can sell a used 1070 for the price of a new 1080 right now.