Ethereum GPU mining?

[codyray10]

^how do you actually get that low? I can't go below 1250 mem on my 290s using afterburner.

Have you tried undervolting in AB at all?

 

[Mercennarius]

I'm not 100% sure I understand the soft fork vs hard fork but from what I think I understand the hard fork seems the most logical to me. The hard fork would reverse the wrong doing and at the same time prevent that exploit from happening again by fixing the system code. The soft fork does neither of these?

 

[n0x1ous]

I'm not 100% sure I understand the soft fork vs hard fork but from what I think I understand the hard fork seems the most logical to me. The hard fork would reverse the wrong doing and at the same time prevent that exploit from happening again by fixing the system code. The soft fork does neither of these?

I'm in the same boat. If someone knows could you explain the differences and consequences of hard vs soft fork?

EDIT: dwarfpool just added softfork voting.

 

[poofyhairguy]

I'm in the same boat. If someone knows could you explain the differences and consequences of hard vs soft fork?

Sure. As I best understand it:

Soft Fork- The Dark DAO is basically blacklisted and cannot trade with any other addresses (except maybe a Robin Hood DAO that is whitelisted to attack it and take the ETH back). This is accomplished via updates to geth and wallet software so that way its a pretty painless transition. Most of the opposition are people who want to do nothing because they love Bitcoin's anarchy

Hard Fork- A whole new blockchain is created and miners move to that blockchain instead of the old one. In that blockchain the old DAO doesn't exist anymore and is turned into a simple contract by brute force, and all the ETH in the dark DAO is moved to this simple contract by brute force. A hard fork is basically what will happen when we move to POS- the old blockchain (with the normal DAO and dark DAO) will still exist until 51% of miners move to the new one and its not viable anymore. Most of the opposition are people who feel it sets a bad precedent to rollback a transaction.



Personally I don't think the hard fork is needed. I think the soft fork will do enough to get most of the DAO back, it just might take months instead of overnight as the locked up dark DAO is slowly drained hack by hack. Seeing as how the hard fork could really split the community (ESPECIALLY if the owners of the Dark DAO bribes miners to continue on the old chain like has been suggested) its a big risk, while a soft fork would be pretty seamless. That is all my opinion though.

 

[poofyhairguy]

I'm not 100% sure I understand the soft fork vs hard fork but from what I think I understand the hard fork seems the most logical to me. The hard fork would reverse the wrong doing and at the same time prevent that exploit from happening again by fixing the system code. The soft fork does neither of these?

Nah a hard fork doesn't prevent anything. The Ethereum system code was fine, the problem was the DAO code. (Actually really the problem is that the DAO got too big and its failure risks everything they have planned for POS, but that is another story)

The only thing that prevents this from happening again is the good sense of people learning from experience not to trust any old smart contract just because the website looks cool.

 

[Mercennarius]

Sure. As I best understand it:

Soft Fork- The Dark DAO is basically blacklisted and cannot trade with any other addresses (except maybe a Robin Hood DAO that is whitelisted to attack it and take the ETH back). This is accomplished via updates to geth and wallet software so that way its a pretty painless transition. Most of the opposition are people who want to do nothing because they love Bitcoin's anarchy

Hard Fork- A whole new blockchain is created and miners move to that blockchain instead of the old one. In that blockchain the old DAO doesn't exist anymore and is turned into a simple contract by brute force, and all the ETH in the dark DAO is moved to this simple contract by brute force. A hard fork is basically what will happen when we move to POS- the old blockchain (with the normal DAO and dark DAO) will still exist until 51% of miners move to the new one and its not viable anymore. Most of the opposition are people who feel it sets a bad precedent to rollback a transaction.



Personally I don't think the hard fork is needed. I think the soft fork will do enough to get most of the DAO back, it just might take months instead of overnight as the locked up dark DAO is slowly drained hack by hack. Seeing as how the hard fork could really split the community (ESPECIALLY if the owners of the Dark DAO bribes miners to continue on the old chain like has been suggested) its a big risk, while a soft fork would be pretty seamless. That is all my opinion though.

Thanks.

But does the soft fork prevent this type of occurrence from happening in the future?

 

[n0x1ous]

Sure. As I best understand it:

Soft Fork- The Dark DAO is basically blacklisted and cannot trade with any other addresses (except maybe a Robin Hood DAO that is whitelisted to attack it and take the ETH back). This is accomplished via updates to geth and wallet software so that way its a pretty painless transition. Most of the opposition are people who want to do nothing because they love Bitcoin's anarchy

Hard Fork- A whole new blockchain is created and miners move to that blockchain instead of the old one. In that blockchain the old DAO doesn't exist anymore and is turned into a simple contract by brute force, and all the ETH in the dark DAO is moved to this simple contract by brute force. A hard fork is basically what will happen when we move to POS- the old blockchain (with the normal DAO and dark DAO) will still exist until 51% of miners move to the new one and its not viable anymore. Most of the opposition are people who feel it sets a bad precedent to rollback a transaction.



Personally I don't think the hard fork is needed. I think the soft fork will do enough to get most of the DAO back, it just might take months instead of overnight as the locked up dark DAO is slowly drained hack by hack. Seeing as how the hard fork could really split the community (ESPECIALLY if the owners of the Dark DAO bribes miners to continue on the old chain like has been suggested) its a big risk, while a soft fork would be pretty seamless. That is all my opinion though.

That makes sense, well explained - thanks!

 

[Mercennarius]

A quote from an ex-Ethereum developer:

“So now the money is in a bunch of different places, controlled by a bunch of different people and we are now approaching this point where we can do a minor alteration to the software or it’s going to become essentially a game of I drain your DAO, you drain my DAO so that no one can ever drain the money and this stalemate has occurred and we still aren’t sure how this will be resolved.”
“The question we are faced with is whether we revert or not, should we set it right? I think that the network should set it right. It is such a terrible failure to lose something like ten percent of the ether in existence to one individual that it could potentially threaten the project. I think we are in the first year and it is a reasonable thing to do. It has been done before and hasn’t grossly effected how bitcoin operates.”

http://www.techworld.com/security/e...bes-night-crypto-currency-was-hacked-3642315/

 

[poofyhairguy]

But does the soft fork prevent this type of occurrence from happening in the future?

No, but neither does a hard fork.

True fixes, like ways to protect smart contracts from these attacks, aren't really on the agenda. Putting out the fire is, and then down the road maybe some other fork might address some issues.

 

[n0x1ous]

Seems like eth has bounced back in value to around $14 USD. A good sign.

 

[poofyhairguy]

Seems like eth has bounced back in value to around $14 USD. A good sign.

Yeah the Robin Hood DAO really opened things up. Now the attacker is counter attacking that DAO and its turning into a real fight. Today is easily one of the more entertaining days in Ethereum's history.

 

[SK10H]

Interesting, I can get 25 MH/s @ ~150W with a 290. So 24 MH/s @ ~100W would be an improvement. If there were a better selection of reliable 6-slot PCIe boards out there, mass-buying 480s would be a good play. Makes you wonder what the 470 can do?

Also, nanopool.org is now letting people vote on the soft fork by "voting with their hash power". They are not voting on the hard fork, and the people running nanopool seem hostile to the idea of the hard fork.

Personally I think that if they are going to do any fork at all, they might need to go all the way, but what do I know . . .

So I'm conflicted as to what to do? I certainly don't want to vote "no" and just leave things as they are, but I'd rather support the hard fork? Not sure if they will have another round of voting for that option?

To be democratic, as a pool, they should just let miners decide and stop being the boss, or it's centralized voting. :\

I just vote yes on dwarfpool.

 

[poofyhairguy]

FYI here is another writeup on the options that is pro-Hard Fork so I don't come off as biased:

https://blog.slock.it/white-hat-siphoning-has-occurred-what-now-f7ba2f8d20ef#.fo8xolx10

 

[Madpacket]

There's some unintentional misguided posts here about the fork(s).

Although not very technical I suggest everyone watch this 20 minute talk by Dr.Gavin Wood (co-founder of ethereum) on the subject.

https://www.youtube.com/watch?v=KaOGtH7J0WE

He explains the situation well.

 

[Madpacket]

FYI here is another writeup on the options that is pro-Hard Fork so I don't come off as biased:

https://blog.slock.it/white-hat-siphoning-has-occurred-what-now-f7ba2f8d20ef#.fo8xolx10

Thanks for the link. Sums up the last 24 hours pretty well. Hard fork is the optimal solution.

 

[Madpacket]

I have read somewhere that 1125 is the best for memory on Hawaii to optimize the timings.

There's no single memory speed that's best as it depends on your goals (raw speed vs power efficiency or a balance of the two). It's mostly dependent on the set GPU speed and type and memory timings (to a lesser extent).

I use GPU-Z to monitor the memory bandwidth and tweak accordingly speeds accordingly.

 

[thilanliyan]

Do you have extended overclocking limits enabled?

When I did that last time, my clocks got stuck at very low levels and the only way to reverse it was to do a clean driver install.

I can "see" below 1250 on the slider in AB (even without extending the overclocking limits), but when I choose lower and click "apply" it just goes back to 1250.

 

[poofyhairguy]

There's no single memory speed that's best as it depends on your goals (raw speed vs power efficiency or a balance of the two). It's mostly dependent on the set GPU speed and type and memory timings (to a lesser extent).

I use GPU-Z to monitor the memory bandwidth and tweak accordingly speeds accordingly.

Have you figured out some sort of guidelines you would like to share?

My 390x I run the core at about 1125 and my 290 is running at 1175 but the memory on both is 1125 because I am sucker for believing things I read on the internet.

 

[EightySix Four]

FYI here is another writeup on the options that is pro-Hard Fork so I don't come off as biased:

https://blog.slock.it/white-hat-siphoning-has-occurred-what-now-f7ba2f8d20ef#.fo8xolx10

I don't really feel like Slock.it is unbiased... they made the DAO. Plenty of people on either side that actually don't have a stake in it.

 

[Elixer]

Hmm, http://arstechnica.com/security/201...ghts-for-its-survival-after-50-million-heist/ basically, a flaw in the exchange made it so crooks got 50 million so far...and they can't fix the bug.

So, doesn't that mean that this cryptocurrency is about to take a nosedive?
I mean, even after the hard/soft fork, there are no real ways to protect something like this besides causing lots of chaos.

The hacks pose an existential threat not only to The DAO but to the entire Ethereum currency. To avert the crisis, Ethereum officials have proposed rolling back the blockchain in a way that would invalidate the stolen ether. Such a "soft fork" of the Ethereum protocol would have to be approved by 51 percent of Ethereum miners in the form of a software update they would install on their servers. Ethereum founder Vitalik Buterin has said he supports such a plan, but in the same statement, he also recognized it would have to be supported by a majority of the miners, meaning it's out of his hands.

The proposal has set off howls of dissent among some Ethereum proponents. They point out that Ethereum was designed to work with its own dedicated programming languages that allow it to work seamlessly with "smart contracts." Such software-driven contracts allow for the automatic payment of funds when a set of detailed conditions are met. Fork opponents say the entire appeal of Ethereum is its decentralized nature that by design is supposed to be immune to control by banks, governments, or other powerful groups.

 

[poofyhairguy]

I don't really feel like Slock.it is unbiased... they made the DAO. Plenty of people on either side that actually don't have a stake in it.

Yeah I was more saying I don't want someone to feel I am pushing a no-hard fork agenda on here so I posted the other side.

Honestly that Youtube link is the best source.

 

[poofyhairguy]

basically, a flaw in the exchange made it so crooks got 50 million so far...and they can't fix the bug.

Nah, its just a flaw in the smart contract of the DAO. The technology of Ethereum is fine, just this one contract (made by Slock it and not the Ethereum foundation) was crap.

There is some debate that the abstraction layer itself has a flaw, but that can be fixed.

So, doesn't that mean that this cryptocurrency is about to take a nosedive?

No one knows. It depends on how it is handled and what the market's reaction to that is.

Seeing as how Eth is around $13 in the middle of the crisis I feel good about its future. But Litecoin was hot at one point and cooled, you never know.

 

[EightySix Four]

Nah, its just a flaw in the smart contract of the DAO. The technology of Ethereum is fine, just this one contract (made by Slock it and not the Ethereum foundation) was crap.

There is some debate that the abstraction layer itself has a flaw, but that can be fixed.

To reiterate - no issue with Ethereum or the EVM (ethereum virtual machine). This is an application running on Ethereum that had a flaw, which some are arguing (with fair reasoning) was caused by the language that is used to write ethereum contracts being too high level and abstracting away too many things from the devs. This was done to make it easier to get developers onboarded into the Ethereum ecosystem from similarly simple languages like JS, which in hindsight was a bad idea.

 

[Feld]

This is a really excellently-explained take on why a hard fork should be done:

http://forums.prohashing.com/viewtopic.php?f=11&t=871

 

[Elixer]

If all that is required is 51% for a fork, then it seems to me this can be abused as well.
Yes, a hard fork should be done (don't allow crooks to get away with this crap), but how the heck do you make up lost mining that is currently going on?

It is a lose lose situation here, the longer this goes on, the faster the crook(s) can try and cash out, yet, people are still mining, which will end up being worthless.

Bottom line is, they should have already done the hard fork, and call it lesson learned, and people should really stop mining until this mess is straightened out, since it will end up being worthless until the fork is done.