Pages:
Author

Topic: Last month more GPUs went online than entire combined GPUs during Litecoin days (Read 2997 times)

legendary
Activity: 1498
Merit: 1030
Giving this thread a bump with an update

Last May there was basically more GPUs that went online during the entire combined Litecoin mining days back in late 2013 early 2014.

Back then it was estimated that about 320,000 GPUs were actively mining.

Last May there was only 350,000 GPUs being added monthly.

Now things have basically risen exponentially.

Last month about 2.1 million GPUs were added to the ETH network. About a 6-7x increase.

So basically more GPUs are added every 5 days than the entire Litecoin GPUs in circulation of 2014

More GPUs are added every 15 days than the entire ETH GPUs in circulation when I created this thread back last May.


During the Litecoin days, all AMD GPUs were sold out.
Last Summer all AMD and Nvidia were sold out

Now the capacity has risen exponentially so I am guessing that AMD and Nvidia somehow have increased production in the last year or so. Because too many GPUs are coming online .

I never had an issue buying AMD GPUs during the Litecoin mining days - with NO "limit 1" restrictions ANYWHERE.
I do remember a few reports of specific retailers being out on occasion of specific GPU model cards - usually the 7870.

Nvidia seems to be in major shortage due to a combination of TSMC switching Pascal GPU production over to Volta production "around the end of the year", and a VERY good Christmas sales season.
AMD - partly the Christmas sales season, partly mining demand.

Last SUMMER the supply of GPUS gradually IMPROVED, it was last SPRING that it got really bad for a while with AMD cards pushing into Nvidia 1070 range and even low-end 1080 range, and A FEW Nvidia models having severe supply issues for less than a month and some mild price gouging.
That has NOT happened this time around - though it's been close - and Nvidia is who ran out FIRST.
legendary
Activity: 1498
Merit: 1030
Quote from: adaseb
Its really hard to imagine these record breaking hashrate levels without ASICs in the Ethereum ecosystem.

Take Litecoin for example during the 2013-2014 Bubble.

The hashrate peaked at ~120GH/s on Jan 2014. If you were around during the Scrypt Litecoin days you would know that unlike today there were only 2 coins to mine back then. Litecoin AND Dogecoin. Dogecoin hashrate was around ~100GH/s.

However keep in mind that sometime during the end of 2013, a small era of Scrypt ASICs started emerging. So lets assume that 120GH/s was for SCRYPT GPU mining only. And another 120GH/s was being mined with the newly released 300KH/s scrypt asics.

A 7970/280X mined scrypt at 750KH/s so around ~120GHs/0.75 = ~160,000 GPUs mining. If we assume all those scrypt asics don't count then lets say ~320,000 GPUs mining.

According to etherscan.io last month a total hashrate gain of 9,104GH/s was added. Since each RX 470 ~ 28MH/s that equals to about ~325,142 GPUs. Also another 10% of that ~32,500 most likely went into the ETC hashrate growth. For simplicity I am not taking ZEC, ZCL, XMR into account.

So as much as its fun to ROI your GPU in less than 1 month, I am wondering whats going to happened when the bubble bursts. In Jan 2014, mining was VERY profitable like it is today, and in less than 6 months, it started using more electricity then profits.

However now we have most likely 10x as many miners and GPUs out there. So I can only imagine how horrific it might get when either the BTC bubble bursts, alt-coins bubble burts, ETH lowers block rewards to 3 ETH per block, ETH goes POS at end of year...


There was also DarkCoin (now known as DASH, the original X11 coin) among others - but the others didn't use a ton of GPUs.

The Gridseed ASIC didn't show up 'till early 2014 - which is what drove a ton of GPUs into X11 mining for a few weeks 'till IT overloaded, then drove most of those GPUs out of cryptocoin mining entirely
http://cryptomining-blog.com/1158-first-impressions-from-the-gridseed-5-chip-scrypt-asic/ was one of the first reviews of the original Gridseed Orb - in MAR 2014, indicating they didn't show up 'till Feb sometime.
It also took time for a significant number to deploy - Scrypt would have still; been mostly GPU mined well into March and probably most of April, when the hashrates started climbing FAST on both Litecoin AND Doge.

MOST Scrypt miiners preferred the 7870 to the 7970 or 7950 as they were a ton more COST effective, and a LOT of miners used the 7850 for the same reason.
Half or less the cost for 80% ballpark of the hashrate MATTERED.

The "average" Scrypt mining card was probably a lot closer to 600 than 750, but that would still only be about 200,000 GPUs mining Litecoin ALONE at the peak in January a month before the Gridseed Orb was introduced - and probably about the same number for DOGE at IT'S peak in Feb 2014 pre-Gridseed, so likely a half million or so TOTAL between those two and everything else of the time.

OTHER THAN Ethereum, all of the major coins like ZEC were *FLAT TO DROPPING* a hair on hashrate during January and early Febuary - in the case of ZEC, it was flat to down from a peak in NOVEMBER to today.


The crash this time around isn't going to be as bad though, as there ARE other coins around that can soak all of the hashrate while still staying profitable - as long as PRICES don't collapse in the meantime.
legendary
Activity: 1848
Merit: 1166
My AR-15 ID's itself as a toaster. Want breakfast?

So for me it's not scarey at all - but those people running say 5 or more rigs are just asking to wake up one day and discover they have a lot of hardware and nothing to do with it. I understand the motivation to increase in scale when I see what I have done with just 10-12 GPUs, but that is my limit.

 Worst case, all of the hardware I've bought in the last year for mining *still has good resale value* because it is current or half-generation old.

 If you've been paying attention, RX 4xx series cards are sometimes selling now for MORE than they sold for new, due to the short-term scarcity of the RX 5xx parts as AMD gets those ramped up and the resulting price gouging raising the floor on the value of the older "IN SOME WAYS BETTER but overall generally a tossup" parts.

 Realistically though, if I ended up having to sell in 6 months, I could still probably realise 80% or more of what I paid for my recently purchaced gear - worst case rebuild it into "gaming machines" and might even be able to sell at a small profit.
 Any of the gear I have that was NOT recently purchased is long-since paid for and PURE PROFIT making at this point (except for electric and cooling costs) and most of that gear I'd probably be running anyway on various "distributed network projects" like DNet and F@H.

 POS and the eventual end of ETH POW is not FUD - the most recent statement out of Vitalik is still "aiming for POS end of year or early next year" but planning to impliment in stages, and the first stage is supposed to be ready "soon".

 I won't be shocked if there are further delays though well into next year.

 I still see no significant probability of an ETH ASIC.

 I can see a possibility of the "diff bomb" driving most miners out of ETH before ETH goes full POS, but that's still months away and it will probably get "adjusted" again if the POS work goes slower than currently planned.

You actually think ETH is going to go pure POS? 

Amazing.  I suppose you have a pet unicorn as well.

Believing the ETH devs will create a POS system which works as they claim; is like believing they are going to prove P = NP.

No one needs to worry about an ETH asic not because companies are worried about POS; but because the ETH devs would simply fork to a sufficiently different version of Ethash: which rendered the asics inert.  Diff bomb, please; what a joke.  ETH would be quickly replaced if its devs actually implemented any of this nonsense.  They already have too many problems.


doesn't mean people won't jump on the asics for ETC anyways;  which makes PERFECT sense as to why ETC has jumped up so much in the last little bit since the announcement of the new bitmain asic.
legendary
Activity: 3808
Merit: 1723
Giving this thread a bump with an update

Last May there was basically more GPUs that went online during the entire combined Litecoin mining days back in late 2013 early 2014.

Back then it was estimated that about 320,000 GPUs were actively mining.

Last May there was only 350,000 GPUs being added monthly.

Now things have basically risen exponentially.

Last month about 2.1 million GPUs were added to the ETH network. About a 6-7x increase.

So basically more GPUs are added every 5 days than the entire Litecoin GPUs in circulation of 2014

More GPUs are added every 15 days than the entire ETH GPUs in circulation when I created this thread back last May.


During the Litecoin days, all AMD GPUs were sold out.
Last Summer all AMD and Nvidia were sold out

Now the capacity has risen exponentially so I am guessing that AMD and Nvidia somehow have increased production in the last year or so. Because too many GPUs are coming online .
legendary
Activity: 1498
Merit: 1030
one thing to note though, as long as there are no graphic cards to buy diff generally won't rise dramatically...


 There are cards to buy, just not as many as folks want and at a lot higher prices than most are willing to pay.

 But I do agree that the current AMD card shortage is going to hobble diff increases for a while on many coins.

member
Activity: 130
Merit: 11
one thing to note though, as long as there are no graphic cards to buy diff generally won't rise dramatically...
sr. member
Activity: 612
Merit: 250

So as much as its fun to ROI your GPU in less than 1 month, I am wondering whats going to happened when the bubble bursts.


10x the gear is fully justified.  should be 15x the gear.

the market cap for alt coins is more the 30 billion dollars back in the day ltc was around 2 billion  so we are at least ⅔ of what we would need to be to match the ltc bubble.  which popped due to asic 's coming in and off book mining by zen josh gaw criminals.

I think we are good for a few months at least.

 I don't expect the pricing to burst soon - the movement of capitol into Crypto, coupled with increasing PUBLIC acceptance, seems likely to keep prices up for a while.

 I expect the PROFITABILITY bubble to burst when ETH goes mostly POS - but that's end of year at the earliest, and probably somewhat longer even if they DON'T have any more delays.



I think the phasein of the PoS will take many years. So the mining profitability will just reduce slowly.
legendary
Activity: 1498
Merit: 1030

So as much as its fun to ROI your GPU in less than 1 month, I am wondering whats going to happened when the bubble bursts.


10x the gear is fully justified.  should be 15x the gear.

the market cap for alt coins is more the 30 billion dollars back in the day ltc was around 2 billion  so we are at least ⅔ of what we would need to be to match the ltc bubble.  which popped due to asic 's coming in and off book mining by zen josh gaw criminals.

I think we are good for a few months at least.

 I don't expect the pricing to burst soon - the movement of capitol into Crypto, coupled with increasing PUBLIC acceptance, seems likely to keep prices up for a while.

 I expect the PROFITABILITY bubble to burst when ETH goes mostly POS - but that's end of year at the earliest, and probably somewhat longer even if they DON'T have any more delays.

legendary
Activity: 1848
Merit: 1166
My AR-15 ID's itself as a toaster. Want breakfast?
if you ask me...

"when the bubble pops" I.E. when all the sheep have to move their herd  Roll Eyes ....

Multi algo profit switching will reign far superior profits.  So whomever has the most versatile GPU farm (best overall hashrates in all algos, not just one or two taken into account) will reap the most rewards.


Multi algo mining was most profitable for me when altcoin mining was in its huge slump.   When it started picking up, I started locking down on the most profitable algo for my GPU type/generation at the time.   Which is where i'm still at right now =)
sr. member
Activity: 308
Merit: 250


 Current top miner on flypool added over 200 rigs today, most of THOSE look like single-card NVidia or dual-card RX-470 ballpark....


Most of cracks on TPB install miners these days.
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
Its really hard to imagine these record breaking hashrate levels without ASICs in the Ethereum ecosystem.

Take Litecoin for example during the 2013-2014 Bubble.

The hashrate peaked at ~120GH/s on Jan 2014. If you were around during the Scrypt Litecoin days you would know that unlike today there were only 2 coins to mine back then. Litecoin AND Dogecoin. Dogecoin hashrate was around ~100GH/s.

However keep in mind that sometime during the end of 2013, a small era of Scrypt ASICs started emerging. So lets assume that 120GH/s was for SCRYPT GPU mining only. And another 120GH/s was being mined with the newly released 300KH/s scrypt asics.

A 7970/280X mined scrypt at 750KH/s so around ~120GHs/0.75 = ~160,000 GPUs mining. If we assume all those scrypt asics don't count then lets say ~320,000 GPUs mining.


According to etherscan.io last month a total hashrate gain of 9,104GH/s was added. Since each RX 470 ~ 28MH/s that equals to about ~325,142 GPUs. Also another 10% of that ~32,500 most likely went into the ETC hashrate growth. For simplicity I am not taking ZEC, ZCL, XMR into account.

So as much as its fun to ROI your GPU in less than 1 month, I am wondering whats going to happened when the bubble bursts. In Jan 2014, mining was VERY profitable like it is today, and in less than 6 months, it started using more electricity then profits.

However now we have most likely 10x as many miners and GPUs out there. So I can only imagine how horrific it might get when either the BTC bubble bursts, alt-coins bubble burts, ETH lowers block rewards to 3 ETH per block, ETH goes POS at end of year...



10x the gear is fully justified.  should be 15x the gear.

the market cap for alt coins is more the 30 billion dollars back in the day ltc was around 2 billion  so we are at least ⅔ of what we would need to be to match the ltc bubble.  which popped due to asic 's coming in and off book mining by zen josh gaw criminals.

I think we are good for a few months at least.
legendary
Activity: 1498
Merit: 1030
has to be a chinese farm, look at how many workers he has , his power usage is probbaly in the 50k per month range


So much ignorance on these forums. 750 workers isnt even 1 megawatt of power, that farm can be anywhere. Please stop spreading falsehoods. Its time people stop saying "China" for everything crypto that is at any sort of scale.


Plus each one of his miners seems to be 1 gpu miners. So likely hundreds of gaming computers setup all around world...maybe

 They do seem to have shifted a lot of their hashrate OUT of ETH the past couple days.
 Probably moved some over to ZEC?

 Current top miner on flypool added over 200 rigs today, most of THOSE look like single-card NVidia or dual-card RX-470 ballpark....
sr. member
Activity: 465
Merit: 252
has to be a chinese farm, look at how many workers he has , his power usage is probbaly in the 50k per month range


So much ignorance on these forums. 750 workers isnt even 1 megawatt of power, that farm can be anywhere. Please stop spreading falsehoods. Its time people stop saying "China" for everything crypto that is at any sort of scale.


Plus each one of his miners seems to be 1 gpu miners. So likely hundreds of gaming computers setup all around world...maybe
sr. member
Activity: 612
Merit: 250
do you think they will actvate pos but keep the pow phase? would be better for other altcoin profit, the nethahs of ethereum would not reverse on other altcoin, but i see no news about this possibility just rumors, also i don't the reaosn behind going pur epos is because they fear asic that is just what you believe

Initially, there will be PoW+PoS, then more PoS over the years.
sr. member
Activity: 406
Merit: 250

So for me it's not scarey at all - but those people running say 5 or more rigs are just asking to wake up one day and discover they have a lot of hardware and nothing to do with it. I understand the motivation to increase in scale when I see what I have done with just 10-12 GPUs, but that is my limit.

 Worst case, all of the hardware I've bought in the last year for mining *still has good resale value* because it is current or half-generation old.

 If you've been paying attention, RX 4xx series cards are sometimes selling now for MORE than they sold for new, due to the short-term scarcity of the RX 5xx parts as AMD gets those ramped up and the resulting price gouging raising the floor on the value of the older "IN SOME WAYS BETTER but overall generally a tossup" parts.

 Realistically though, if I ended up having to sell in 6 months, I could still probably realise 80% or more of what I paid for my recently purchaced gear - worst case rebuild it into "gaming machines" and might even be able to sell at a small profit.
 Any of the gear I have that was NOT recently purchased is long-since paid for and PURE PROFIT making at this point (except for electric and cooling costs) and most of that gear I'd probably be running anyway on various "distributed network projects" like DNet and F@H.

 POS and the eventual end of ETH POW is not FUD - the most recent statement out of Vitalik is still "aiming for POS end of year or early next year" but planning to impliment in stages, and the first stage is supposed to be ready "soon".

 I won't be shocked if there are further delays though well into next year.

 I still see no significant probability of an ETH ASIC.

 I can see a possibility of the "diff bomb" driving most miners out of ETH before ETH goes full POS, but that's still months away and it will probably get "adjusted" again if the POS work goes slower than currently planned.

You actually think ETH is going to go pure POS?  

Amazing.  I suppose you have a pet unicorn as well.

Believing the ETH devs will create a POS system which works as they claim; is like believing they are going to prove P = NP.

No one needs to worry about an ETH asic not because companies are worried about POS; but because the ETH devs would simply fork to a sufficiently different version of Ethash: which rendered the asics inert.  Diff bomb, please; what a joke.  ETH would be quickly replaced if its devs actually implemented any of this nonsense.  They already have too many problems.



do you think they will actvate pos but keep the pow phase? would be better for other altcoin profit, the nethahs of ethereum would not reverse on other altcoin, but i see no news about this possibility just rumors, also i don't the reaosn behind going pur epos is because they fear asic that is just what you believe
legendary
Activity: 2294
Merit: 1182
Now the money is free, and so the people will be
companies like pandaminer are to blame also, they make large scale GPU operations so much easier to set up.  I'm still going to mine (at least in winter) regardless of power costs or profit - I have to pay for the heat in my house anyway.  Large scale farms with rent, employees, power bills, property taxes (maybe), etc will have more trouble staying alive.  They'll dump their gear, which will probably be at the end of its lifetime due to being run hard and hot, and you'll get cheap, cheap cheap GPU's on ebay.  But when ?  well after the current btc bubble run.  which will be epic.  could be years.
newbie
Activity: 32
Merit: 0
Food for thought!

It's a certain that at one point in near future eth, zcash, monero won't be profitable to mine. The way how I am looking at it. Once miners (rigs) get their ROI, ppl will just mine new coin and horde them.
Most of the miners that I talked to no one is really afraid to lose what they have put in.
Welp enjoy the money while it lasts:)
member
Activity: 76
Merit: 10
Geniuses are always branded as crazy

So for me it's not scarey at all - but those people running say 5 or more rigs are just asking to wake up one day and discover they have a lot of hardware and nothing to do with it. I understand the motivation to increase in scale when I see what I have done with just 10-12 GPUs, but that is my limit.

 Worst case, all of the hardware I've bought in the last year for mining *still has good resale value* because it is current or half-generation old.

 If you've been paying attention, RX 4xx series cards are sometimes selling now for MORE than they sold for new, due to the short-term scarcity of the RX 5xx parts as AMD gets those ramped up and the resulting price gouging raising the floor on the value of the older "IN SOME WAYS BETTER but overall generally a tossup" parts.

 Realistically though, if I ended up having to sell in 6 months, I could still probably realise 80% or more of what I paid for my recently purchaced gear - worst case rebuild it into "gaming machines" and might even be able to sell at a small profit.
 Any of the gear I have that was NOT recently purchased is long-since paid for and PURE PROFIT making at this point (except for electric and cooling costs) and most of that gear I'd probably be running anyway on various "distributed network projects" like DNet and F@H.

 POS and the eventual end of ETH POW is not FUD - the most recent statement out of Vitalik is still "aiming for POS end of year or early next year" but planning to impliment in stages, and the first stage is supposed to be ready "soon".

 I won't be shocked if there are further delays though well into next year.

 I still see no significant probability of an ETH ASIC.

 I can see a possibility of the "diff bomb" driving most miners out of ETH before ETH goes full POS, but that's still months away and it will probably get "adjusted" again if the POS work goes slower than currently planned.

You actually think ETH is going to go pure POS? 

Amazing.  I suppose you have a pet unicorn as well.

Believing the ETH devs will create a POS system which works as they claim; is like believing they are going to prove P = NP.

No one needs to worry about an ETH asic not because companies are worried about POS; but because the ETH devs would simply fork to a sufficiently different version of Ethash: which rendered the asics inert.  Diff bomb, please; what a joke.  ETH would be quickly replaced if its devs actually implemented any of this nonsense.  They already have too many problems.



https://etherscan.io/chart/blocktime you can see it slowly increasing
member
Activity: 96
Merit: 10

 Worst case, all of the hardware I've bought in the last year for mining *still has good resale value* because it is current or half-generation old.

Or worst case is all the hardware you bought isn't worth anything because 50,000 video cards go up on ebay  Cheesy
legendary
Activity: 2408
Merit: 1102
Leading Crypto Sports Betting & Casino Platform
has to be a chinese farm, look at how many workers he has , his power usage is probbaly in the 50k per month range


So much ignorance on these forums. 750 workers isnt even 1 megawatt of power, that farm can be anywhere. Please stop spreading falsehoods. Its time people stop saying "China" for everything crypto that is at any sort of scale.

50 bucks says its china  Grin Grin
Pages:
Jump to: