Ethereum GPU mining?

Page 209 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Oct 14, 2003
5,854
105
126
For ETH I don`t know, they have tech that but not the speed needed for smart contracts, also ETH development is 100% centralized and no one believe in their chain which is controlled by one man
So you are driven by ideals, and not reality then. If you don't have a leader at all you end up with a situation where people argue and never goes anywhere. It especially doesn't help when the market is immature as cryptos are.

I flat out reject the decentralization-for-the-sake-of-decentralizing idea embraced by many in the crypto community.

Some people look up to crypto for the solution to mankind's problems. No, its a tool. Your life doesn't revolve around the tool. A better tool augments your skill. Just like any tools though, it can be used against society.

It's never about the tech. It's always about the people.
 
Apr 27, 2000
11,018
642
126
It's never about the tech. It's always about the people.
In the case of Bitcoin, they have lousy tech and lousy people. They lose on both counts.

The EF better hit the turbo boost on development since they are getting left behind. There's still good people in there, but without the tech, they will lose out to other blockchain projects.
 
Dec 31, 2009
129
4
76
Just watch BTC survive and ICO and tokens drop dead... they are worthless without of decentralization
ETH need to do few big steps (decentralize) which they don`t like or it will die one day, also ETH tech is pure REDACTED talk, they delivered zero promises over the years only reason to hold was greedy miner

About the roger ver ship, just compare it with the last ETH one, both camps are insane and full of morons

We have a zero tolerance policy about profanity in the tech sub-forums.
Let this be the last one you do.

Iron Woode

Super Moderator
 
Last edited by a moderator:
Apr 27, 2000
11,018
642
126
Just keep preachin brotha. I'm sure we'll all bow to BTC sooner or later. Or not.
 
Oct 14, 2003
5,854
105
126
The EF better hit the turbo boost on development since they are getting left behind. There's still good people in there, but without the tech, they will lose out to other blockchain projects.
The size can be seen as an advantage because there are many more resources out there and community to help you if you want to start in dapp development. Or wallets and UI are usually better for the end user. It's also much faster to develop if you start from scratch, but you have no one on it.

Would Constantinople allow for the full functionality of Plasma?
Plasma, is different from state channels. Here's a nice article about the differences between the two: https://blog.ethlend.io/journey-to-...e-channels-plasma-and-plasma-cash-4fe7914e06a

(Plasma by the way is developed by others. Notably the Loom Network)

Here's an article about counterfactual state channels: https://medium.com/coinmonks/unders...ment-channels-and-state-channels-9e939d7c6f34

Here's one about EIP1014, which will be included in Constantinople: https://blog.cotten.io/ethereums-eip-1014-create-2-d17b1a184498

From what I understand, not only the EIP will ease the development for dapps using such state channels, but because its a feature of the hard fork it has a chance to be widely used. It probably won't have a huge effect on existing user base but its an incentive for new developers to use Ethereum and maybe offers an upgrade path for current dapps.
 

Headfoot

Diamond Member
Feb 28, 2008
4,388
40
126
The energy usage of Bitcoin is a fundamental problem, it will never truly succeed at scale until that is solved. The technology exists now thanks to new cryptos, but Bitcoin's governance isn't gonna let that happen. Too much vested interest by miners having invested in hardware.
 
Apr 27, 2000
11,018
642
126
The size can be seen as an advantage because there are many more resources out there and community to help you if you want to start in dapp development. Or wallets and UI are usually better for the end user. It's also much faster to develop if you start from scratch, but you have no one on it.
Maybe Ethereum is suffering from its own inertia. Not so much as older projects (see above post by @Headfoot ). Fortunately, Ethereum's miner community has gone along with all the hard forks pretty willingly, outside of that one thing back in 2016 . . .

Plasma, is different from state channels. Here's a nice article about the differences between the two: https://blog.ethlend.io/journey-to-...e-channels-plasma-and-plasma-cash-4fe7914e06a

(Plasma by the way is developed by others. Notably the Loom Network)

Here's an article about counterfactual state channels: https://medium.com/coinmonks/unders...ment-channels-and-state-channels-9e939d7c6f34

Here's one about EIP1014, which will be included in Constantinople: https://blog.cotten.io/ethereums-eip-1014-create-2-d17b1a184498

From what I understand, not only the EIP will ease the development for dapps using such state channels, but because its a feature of the hard fork it has a chance to be widely used. It probably won't have a huge effect on existing user base but its an incentive for new developers to use Ethereum and maybe offers an upgrade path for current dapps.
Looks like I've got some more reading to do then. Thanks!
 
Jul 1, 2001
20,769
85
126
Ya know, if you're sick of losing money mining Ethereum and still want to put your video cards to work, the distributed computing guys are looking for help with their projects.
 
Oct 14, 2003
5,854
105
126
There are dapps out there that continues coming with improved versions. Except in crypto, you can't delete the older versions. So the older versions exist with the newest and the state size keeps increasing. That means the storage and syncing times continue to increase, and some may be not used at all.

Imagine what's going to happen in 20, 30 years if it goes like this! Very few nodes are going to be able to host this.

This is both an advantage and an achilles heel for immutability. Certain solutions to it were suggested. Like rent for active dapps.

Rent could theoretically allow a near-full state sync that's much smaller due to only needing to sync active dapps. Of course I think the costs for rent has to be low.

Edit: I want to have an Optane SSD just to see how syncing will work. I'm thinking since its so demanding on the endurance it likely consists of a lot of random 4K reads/writes. The main thing is the durability though. 5.11PB on just the 280GB 900P is enough to make it last 25 years at 15TB/month.

Yes, its not cheap. Maybe with PoS I'll consider it since there's an incentive. Nodes are necessary for the network to be up.
 
Last edited:
Apr 27, 2000
11,018
642
126
Or you could have routine "sidestep" hard or soft forks intended to orphan certain blocks.
 
Oct 14, 2003
5,854
105
126
Or you could have routine "sidestep" hard or soft forks intended to orphan certain blocks.
IMO this will not happen, because it goes against the ideals, which is decentralization and the immutability it brings. A common example given is an application or a game that's available even if the development stops. But if rent is low, the developer can choose to keep otherwise an obsolete application active.

The algorithm will need to be smart enough to know whether its completely useless or not, and that's nearly impossible.

Interestingly, the results provided by the rent scheme may result in mirroring the outside world in some ways, since it might dynamically disable outdated ones not paying the rent. These things really drive points like "nothing is free" home.
 

Feld

Senior member
Aug 6, 2015
229
13
71
For anyone interested, a new version of geth was released today that the developers say is 15% faster at syncing from scratch, 25% faster at maintaining sync after a full sync has been reached, and fixes various bugs including ones that impaired connectivity to other nodes for syncing.
 
Apr 27, 2000
11,018
642
126
IMO this will not happen, because it goes against the ideals, which is decentralization and the immutability it brings. A common example given is an application or a game that's available even if the development stops. But if rent is low, the developer can choose to keep otherwise an obsolete application active.

The algorithm will need to be smart enough to know whether its completely useless or not, and that's nearly impossible.
I'm not saying it's a good idea, but it has been done in the past (DAO). And if the majority of miners/verifiers voted for such a thing, it would happen, assuming there were EIPs drafted for it. Beyond that, there would have to be curation of exactly what was and was not removed from the chain.

I do agree that if loading a dApp onto the blockchain remains "free" for the most part, that adding chaindata becomes trivially simple, to the point that an attacker could bloat the blockchain to the point of uselessness with very little cost/effort.
 
Oct 14, 2003
5,854
105
126
Beyond that, there would have to be curation of exactly what was and was not removed from the chain.
This is the difficult part though. Unless they want to hire 20 thousand people to look at every application to see what to delete, its going to have to be automated. And automated systems are stupid.
 
Apr 27, 2000
11,018
642
126
This is the difficult part though. Unless they want to hire 20 thousand people to look at every application to see what to delete, its going to have to be automated. And automated systems are stupid.
Late reply, but I think anyone that puts a dApp on the blockchain ought to be able to include kill code or replacement code to handle version upgrades. That would be easy to exploit in the case of scammy dApps/smart contracts though. Convince people to load ETH into the contract, then pull the plug and kill it after absconding with the tokens.

In other news, ETH temporarily crashed to $13 on Coinbase Pro. Hope you guys were posted up to take advantage of that (I wasn't; too bad!). The price still hasn't recovered fully, and the rest of the market chased ETH downwards. Pretty funny stuff.
 

IntelUser2000

Elite Member
Oct 14, 2003
5,854
105
126
Bad actors will always exist.

but I think anyone that puts a dApp on the blockchain ought to be able to include kill code or replacement code to handle version upgrades.
Some plan and do better, and others will just dive head first and learn as they go. Plus, if they were to rely on this system to reduce state sizes, we have to expect a significant majority to do this, which may not be realistic.

It also veers towards centralized systems. A careful balance is needed between the two. As you said, since so many dapps are financially motivated, easy deactivation will result in even more scams popping up. We may have enough of that already.

Archival nodes can't do anything about this. Because the idea is to keep everything done since the beginning at all times.

In other news, ETH temporarily crashed to $13 on Coinbase Pro.
I hope it doesn't crash to the level where important developers have to leave. If crypto really goes down that low, they might as well not exist.
 
Apr 27, 2000
11,018
642
126
I think some devs are already on the way out the door. Frankly, anyone that needed ETH to trade at $400-$1400 to do any work on it, are not going to be the major contributors to the blockchain anyway. Major work was being done in 2016 when it traded at $12 or less. Why does it need to be worth so much?

When I "bought in" I did not think anyone would pay more than $40 for a token until PoS was fully functioning. Oh how things have changed.
 
Dec 31, 2009
129
4
76
POS was never secure nor had a working concept, will sharding fix POS and when is better question to start with....... (if that ever happen)
The price of ETH was inflated by worthless ICO and tokens with zero use, with that I agree with you ETH should never been over 100$ bit, for that you can blame ETH team which did that on purpose
 
Apr 27, 2000
11,018
642
126
you can blame ETH team which did that on purpose
Garbage. Sorry, you've repeated this screed before, and it isn't remotely believable. Hell Buterin tried to stop the RDN ICO by offering a bunch of ERC20s from other projects (Kyber and something else) to bribe the RDN team into halting the ICO. It didn't work and the ICO went through anyway. Buterin openly questioned the insane value run-up in late 2017. Most of the EF has consulted on only a handful of early, high-profile ICO teams.

You would have us believe that Buterin is a scammer, which he is not. And how can you possibly assess whether or not PoS is or ever will be secure? Give me a break.
 
May 24, 2003
50,451
316
126
www.uovalor.com
Damn eth keeps going down fairly steadily. $119 now lol. Oddly enough though I noticed my mining earnings (in eth) are slightly going up. Not sure if it's because the pool I'm with made changes, or if it's some other effect.

GTX 1070 TI's are actually coming down in price on Ebay so it's almost tempting to expand my rig, though I don't know how to tell what's real and what's a scam, I already got caught once. At this point I don't think I'm even making enough to cover hydro costs though so I should maybe perhaps consider calling it quits. I could convert the rig into a gaming machine instead. On the other hand as long as I don't cash in I'm not really losing anything. I think the ethereum tech itself is really interesting and has lot of potential, it just seems nobody has really used it yet. The minute a major organization like a government decides to adopt it in one way or the other, it will probably affect the price positively. Of course trying to speculate stuff like that is pretty much a gamble. Not any different than the stock market.
 
Apr 27, 2000
11,018
642
126
If profits are going up, it's entirely due to hash rate going down. ETH is at $90 now and might drop more.
 
Dec 31, 2009
129
4
76
Yeah Vitalik team did it, they made most tokens and ICO themselfs, but when Vitalik later saw what they/he did he tried to distance himself
ETH for me is dead at the date they decided to edit ETH blockchain (The DAO) and save their own ass... its the first rule you can never broke and they did it

But its shame as a lot of tech have some promising future in it, well maybe one day we see some better project which have no centralized ppl behind to instamine 70% ICO and make their own rules
 
Oct 14, 2003
5,854
105
126
ETH for me is dead at the date they decided to edit ETH blockchain (The DAO) and save their own ass... its the first rule you can never broke and they did it
Codes are inherently buggy. You get that? Every code has bugs. Some are significant, some are not. Some get revealed, some don't.

DAO hack resulted in people losing money. They decided to return the lost money. Is that really a bad thing? Being inflexible and strictly adhering to rules is what stupid robots and AI do.

No matter what people believe crypto will never allow society to be automated, and eliminate human flaws. Such ideals are not born in reality and for most people that type of thinking is abandoned once they reach adulthood.

Cryptos, just like every other tech invented are tools. Tools don't exist to replace you. They are there to serve and enhance you. But they are very much subject to our flaws. If you make crypto to automate things, you shift the blame from individuals to coders. Arguably, that's even worse.
 
Apr 27, 2000
11,018
642
126
Also, the DAO hack was as much a bug in Solidity as it was anything else.
 


ASK THE COMMUNITY

TRENDING THREADS