Ethereum GPU mining?

[Feld]

Yeah, but those 105c temps... How long is this card going to last mining around the clock?

Where are you seeing 105c temps when mining?

 

[Sonikku]

That's how hot the Radeon 7 is said to get under load. Though I suppose you could undervolt for mining. Never the less, it is likely going to be quite the scorcher compared to most GPUs. It is for that reason that it would not surprise me if it died sooner than most cards if operated continuously in the same fashion.

 

[Feld]

That's how hot the Radeon 7 is said to get under load. Though I suppose you could undervolt for mining. Never the less, it is likely going to be quite the scorcher compared to most GPUs. It is for that reason that it would not surprise me if it died sooner than most cards if operated continuously in the same fashion.

I haven't read every review out there, but the few I did read had the Radeon VII peaking at about 85c in stress tests (though even that would be too high for 24/7 mining and would need to be dialed back). If someone is seeing 105c then that is unacceptable, but I don't think that's what people should be expecting based on what I've seen.

 

[Sonikku]

I haven't read every review out there, but the few I did read had the Radeon VII peaking at about 85c in stress tests (though even that would be too high for 24/7 mining and would need to be dialed back). If someone is seeing 105c then that is unacceptable, but I don't think that's what people should be expecting based on what I've seen.

I forgot where I heard 105c now, I think youtube. But here is an excerpt from Gamespot's review where they were running 104c. Quite a long shot from 85c.

"Of all five cards in our review, the Radeon VII unfortunately takes the cake when it comes to load temperature. When accounting for Radeon VII's high TDP, it would be expected for it to run hot, but with a whopping peak load temperature of 104 C, it surpasses every other card by a longshot. Thankfully, this didn't result in any functional issues, but those who want to run a cool system should take note. Radeon VII's cooling solution may not the best for keeping temperatures down, but its fans stay fairly quiet. "

https://www.gamespot.com/articles/radeon-vii-review-can-amds-new-card-handle-4k-pc-g/1100-6464872/

 

[thecoolnessrune]

I forgot where I heard 105c now, I think youtube. But here is an excerpt from Gamespot's review where they were running 104c. Quite a long shot from 85c.

"Of all five cards in our review, the Radeon VII unfortunately takes the cake when it comes to load temperature. When accounting for Radeon VII's high TDP, it would be expected for it to run hot, but with a whopping peak load temperature of 104 C, it surpasses every other card by a longshot. Thankfully, this didn't result in any functional issues, but those who want to run a cool system should take note. Radeon VII's cooling solution may not the best for keeping temperatures down, but its fans stay fairly quiet. "

https://www.gamespot.com/articles/radeon-vii-review-can-amds-new-card-handle-4k-pc-g/1100-6464872/

Yeah there's something very wrong with that setup. It is a statistical outlier compared to almost every other review out there and they should have followed up with AMD to see what was wrong. 115 is the junction point and I think it's obvious card shouldn't get anywhere near that during normal operation.

 

[VirtualLarry]

Anyone mining Grin?

 

[DrMrLordX]

Yeah, but those 105c temps... How long is this card going to last mining around the clock?

See @thecoolnessrune above, and also keep in mind that undervolting is basically mandatory for mining cards anyway. On those rare occasions when I still mine with my VegaFE, I run it at -20% power and undervolted, for 40 MH/s with temps that never go above 65C. Go blower fan go, whooosh. I would not bother mining with a Radeon VII until Wattman works properly or until someone updates tools like OverdriveNTool to tweak voltage and fan profiles.

And even then, 90 MH/s is still not a lot of money mining ETH right now.

Anyone mining Grin?

Not I. Tell us all about it, though.

 

[maddie]

That's how hot the Radeon 7 is said to get under load. Though I suppose you could undervolt for mining. Never the less, it is likely going to be quite the scorcher compared to most GPUs. It is for that reason that it would not surprise me if it died sooner than most cards if operated continuously in the same fashion.

This might be mistaken.

Previously temps were taken at the die edge, now they're taken at many distinct points within the die, with the highest temp being reported and used as the controlling temp for the regulatory algorithm.

I think that this 105 C temp has always existed but corresponded to a 85-90 C at the edge when measured there.

 

[Red Squirrel]

Still mining eth with my 2 1070's. Got a wooping 1.5879 eth so far which is about $248.57 as I type this. I have not actually sad down to calculate it but I imagine it might be JUST covering hydro costs. Definitly did not pay off even 1 card. They were like $800. They're still going for like $300-$500 now on ebay it's kind of crazy.

Going to keep going for now. Eventually repurpose the rig for something else. Though it is tempting to sell one card while they are still expensive, and then just keep the other for my gaming machine. Been wanting to experiment with a different VM solution than vmware and I believe the cpu I'm using has VT-D so it would make a decent VM server.

 

[DrMrLordX]

Still mining eth with my 2 1070's. Got a wooping 1.5879 eth so far which is about $248.57 as I type this. I have not actually sad down to calculate it but I imagine it might be JUST covering hydro costs. Definitly did not pay off even 1 card. They were like $800. They're still going for like $300-$500 now on ebay it's kind of crazy.

Going to keep going for now. Eventually repurpose the rig for something else. Though it is tempting to sell one card while they are still expensive, and then just keep the other for my gaming machine. Been wanting to experiment with a different VM solution than vmware and I believe the cpu I'm using has VT-D so it would make a decent VM server.

Sorry it didn't work out better for you. On the plus side, that 1.5879 ETH might be worth more in a year's time. I'd hold on to it for a minute to see how that goes. I don't think it's going to go down any further. In fact the whole market is jumping over the last 24 hours so who know's what's going on there?

 

[Red Squirrel]

Sorry it didn't work out better for you. On the plus side, that 1.5879 ETH might be worth more in a year's time. I'd hold on to it for a minute to see how that goes. I don't think it's going to go down any further. In fact the whole market is jumping over the last 24 hours so who know's what's going on there?

Yeah I got in late so it's too be expected. I'll definitely hold on to it for now.

 

[Markfw]

If mining has hurt so many people, and the video card market has died for them, why are people still mining ?

 

[Sonikku]

why are people still mining ?

Because it's what me pappy did and his pappy before em'.

 

[DrMrLordX]

If mining has hurt so many people, and the video card market has died for them, why are people still mining ?

The thing to remember is that if you mine even through unprofitable periods, you can still make bank later when your tokens go up so long as you don't cash them in immediately. I mined a bunch of ETH during the 2016 DAO crash and made a big profit later when it skyrocketed beyond all expectation. Anyone who kept mining through the crypto crash of 2018 saw total token returns go up due to falling hashrates even if profits indexed in USD went down thanks to lowering prices.

Because it's what me pappy did and his pappy before em'.

Oh my.

 

[Feld]

If mining has hurt so many people, and the video card market has died for them, why are people still mining ?

I know you're not a fan of mining, but that's a bit hyperbolic to suggest mining has hurt many people. Mining serves a useful purpose, for now at least, for a technology that many believe will revolutionize large aspects of society once it has matured a little bit more. It's part of the process of improvement.

And it's way less hurtful than, say, somebody buying up GPUs simply to flip them at inflated prices when availability is low and adding zero value. For one thing, blockchain technology like Ethereum has the potential to eliminate parasitic middlemen like this all over, helping society.

 

[Arkaign]

Blockchain is good.

Mining via GPU is fantastically stupid if the processes don't actually produce anything of value intrinsically. Those that leverage for protein folding etc are exceptions.

It's a well understood issue with power waste and is poorly timed considering fossiil fuel issues, climate change, infrastructure strain, etc.

It is understood that the goal is PoW, which should alleviate much of this, or move to ASICs that are an order of magnitude more efficient, but the interim has been pretty bad for most everyone that hasn't seen direct profits from this stage. It distresses me that so much computing power has been dedicated to basically nonsense, when at the very least it could have been harnessed to do beneficial medical or scientific research to 'mine' instead. And one cannot compare mining to gaming. A gamer might average a few hours a day at most, and probably the majority less than that, compared to miners running 24/7/365, sometimes with dozens, hundreds, or even many thousands+ of GPUs. All essentially running like power viruses solving worthless math (outside of the blockchain maintenance itself, a pittance by contrast).

 

[DrMrLordX]

It is understood that the goal is PoW, which should alleviate much of this,

PoS (Proof of Stake) actually, but I get where you're coming from.

 

[Arkaign]

PoS (Proof of Stake) actually, but I get where you're coming from.

Good catch, got my acronyms backwards lol

I don't really fault miners per se, I mean the profits were definitely undeniable for some pretty serious chunks of time. It's just the big picture that turned out to be a bit of a shame. Regardless, blockchain hopefully will dramatically outlive conventional extreme power intensive mining (or algorithms are tied in major crypto towards beneficial work).

 

[Feld]

Blockchain is good.

Mining via GPU is fantastically stupid if the processes don't actually produce anything of value intrinsically. Those that leverage for protein folding etc are exceptions.

It's a well understood issue with power waste and is poorly timed considering fossiil fuel issues, climate change, infrastructure strain, etc.

It is understood that the goal is PoW, which should alleviate much of this, or move to ASICs that are an order of magnitude more efficient, but the interim has been pretty bad for most everyone that hasn't seen direct profits from this stage. It distresses me that so much computing power has been dedicated to basically nonsense, when at the very least it could have been harnessed to do beneficial medical or scientific research to 'mine' instead. And one cannot compare mining to gaming. A gamer might average a few hours a day at most, and probably the majority less than that, compared to miners running 24/7/365, sometimes with dozens, hundreds, or even many thousands+ of GPUs. All essentially running like power viruses solving worthless math (outside of the blockchain maintenance itself, a pittance by contrast).

Providing security for the blockchain you're mining on is intrinsically valuable. It isn't just in theoretical stages - Ethereum is running numerous valuable and innovative applications today, even though it still needs future scaling upgrades for most of them to really go mainstream.

But ASICs being more energy efficient compared to GPUs doesn't in any way mean a blockchain network running on them will consume less electricity. The security of a blockchain comes down to the cost of equipment and electricity to 51% attack it, not the raw hashrate. If you take a blockchain that is 100% GPU mined and in economic equilibrium, and then introduce ASICs that are 10x more energy efficient than GPUs, all that's going to happen over time is all the GPU miners will be driven out and 10x more ASICs will end up mining on the network than GPUs that had been, so energy consumption is the same as it was before, because that is the level of energy consumption where the economics of mining are in equilibrium.

Ultimately the only solution to cutting the wasteful power consumption is moving away from proof of work entirely - ASICs are just an unnecessary and wasteful step in the middle since they have no other use if mining becomes obsolete. Proof of stake is indeed the answer if it's done right, and the sooner it arrives for Ethereum the better. But it is worth it to have proof of work in the meantime, because developers need to have a foundation to build and learn from over time.

I do get where you're coming from though. In the era of undeniable climate change, energy usage is important. I am a miner because I believe in the technology and want to be a part of bringing about the future it can create, but I only mine at a small scale. And only because I have 100% wind power to run it on. I wouldn't be mining if I was using fossil fuel energy, just for ethical reasons.

 

[DrMrLordX]

Sometimes I still wonder why they never tried doing more to limit hashrate per user or hashrate per address. Some kind of a privatized KYC system along with random blockchain rewards that don't necessarily reward you for having more hashrate on the blockchain.

 

[Arkaign]

Sometimes I still wonder why they never tried doing more to limit hashrate per user or hashrate per address. Some kind of a privatized KYC system along with random blockchain rewards that don't necessarily reward you for having more hashrate on the blockchain.

That would seem a good idea to aid in the preservation of the relative democratization of finance/control as it were. I read a great article about how Bitcoin went from this great idea that brought all this new power to the masses, yet the aggressive reality of major players has disrupted that and led to a handful of megapowered interests that can throw their weight around like JP Morgan I'm the bad old days. I am not wording it well, but I think you know basically the gist of it already. The tech of blockchain and bank-busting freedom are both incredibly wonderful potential, I guess human nature is always going to be a threat. Where one could profit by stepping on a fellow's face, you will surely find bootprints.

 

[Red Squirrel]

A lot of people who are still mining are also doing it because they already invested the time and money in the hardware. May as well put it to use. I never understood this notion that you have to shut down everything the minute profits are down. You see that even with real mining, like gold, oil etc. You know damn well the product is not always going to be low, it only makes sense to keep operations going until it goes back up. Just don't cash out low.

 

[DrMrLordX]

Just don't cash out low.

That works if you can sustain operations without dipping into mining proceeds. If tokens pay the electric bills then you're sunk.

 

[MrTeal]

A lot of people who are still mining are also doing it because they already invested the time and money in the hardware. May as well put it to use. I never understood this notion that you have to shut down everything the minute profits are down. You see that even with real mining, like gold, oil etc. You know damn well the product is not always going to be low, it only makes sense to keep operations going until it goes back up. Just don't cash out low.

If mining profits go negative, a small scale miner should absolutely turn off their gear. You can still believe in the long term viability of the technology and want to bet on the future of it that's fine, but if it costs you $1 to make 90 cents of a coin, it's better to idle your gear and just buy $1 of that coin.

Real mines can't just flip the switch the same way a crypto setup in someone's garage can. Being small scale doesn't give you many of the advantages of the big mining farms, but it does give a lot of flexibility. If you can't cover the cost of the power on a day by day basis, shut it down and assess the situation. You don't have to get out, but you should definitely look at what you're trying to achieve.

 

[DrMrLordX]

So @Grazick you decided to dislike a post from November 2018. Is there something you wanted to say? This is a discussion forum. If you dislike a post, maybe you should consider talking about it instead of just pushing a button.

I'm all ears.