EIP: Modify block mining to be ASIC resistant.

[pipermerriam]

Disclaimer: My area of expertise does not lend well to me suggesting how to make things ASIC resistant. I hope there are some informed opinions floating around out there who can help fill in the how.

According to "the internet" there is an ASIC based ethereum miner on the horizon.

this may be the original source of that news

If you believe the analysis in the comments on this reddit thread BitMain may already running these miners.

I believe it is the accepted wisdom that ASIC based mining leads to increases centralization when compared to GPU mining.

This leads us to two questions:

1. Should we hard fork to make ASIC mining harder and to demonstrate a willingness to hard fork any future ASIC based ethereum mining.
2. What specifically changes do we make to implement this increased ASIC resistance.

Should we fork?

I propose that people indicate support/opposition with a simple 👍 / 👎 on this main issue. I would prefer this conversation not devolve into deep discussions around this subjective topic so my request is that people refrain from commenting on that specific question here.

How do we implement improved ASIC resistance

This is the primary issue that I think needs to be addressed, after which we can have an informed discussion about whether we should actually do it.

[krtschmr]

@MicahZoltu

Making miners happy is not a primary goal of Ethereum. Its primary goals are censorship resistance and trustless interactions.

thanks for stabbing our back 2 years after we backed ethereum it's time to get rid of us. wohoo! maybe we are all greedy and obviously ASICs are taking our profits. some of us have even more then 1M$ invested into GPU mining farms. of course for own benefit, but on the other side we are all big believers in this technology. Miners are the ones that brings ETH towards PoS. Without us there is literally nothing going on. Better make us happy than angry.

If we are forced to leave to to whatever reason (ETH drops 50$ or Network goes 800 TH/s or whatever) then Ethereum is nothing else then a chinese controlled smart contract chain.

[int03h]

Again - it's like everyone has turrettes. Shouting fork fork fork doesn't produce even one line of code.

What is it that everyone would like to fork?

Switching the algorithm is INSANELY idiotic. Tweaking it even a little could break the whole network and potentially destroy video cards, and not even affect the "ASICS".

The devs are not clairvoyant. Coding something just for the sake of appeasing the masses is a waste of their time and ultimately delays PoS.

Doing nothing requires no code, therefore it is a solution to problem that can't be proven to exist.

Increasing the rules around the DAG file size would punish many and might just make the ASICs "break" but it would probably obsolete 1/4 of the network. Again. is a solution. A supremely stupid one, but a solution nonetheless.

And .. if Bitmain has made something new, then we should all embrace it, believe me AMD and NVIDIA are not going to let some random Chinese crytocoin hardware manufacturer that doesn't own a single Fab out manuever them.

The BTC devs all had their fingers on the scale when it came to mining. They saw the innevitable coming very early on when the first Avalon was put on Gavin's desk at the foundation and they purposefully ignored it. What BTC is now is the result.

Dagger Hashimoto has weathered the storm for many years. Land is in sight with PoS. Burning our boats and jumping ship is kinda pointless now.

This isn't Reddit or Bitcointalk or Twitter, either calm down or write some code and commit it.

[AliAshrafD]

Being a programmer long enough to have some voice, I don't like the 'write code or be silent' slogan some folks are repeating over and over. Code comes after the direction has been set, well, in most cases. As of this thread's request for vote, I say 'nah' for the time being but I think we should wait for the shits to be available and find out in what extends they are truly ASICs and how many manufacturers can compete and keep the game fair enough. For instance, I have this impression: Bitmain has designed an Application Specific System (instead of Integrated Circuit) and it can be easily decoded and replicated by other companies (including small ones). If it turns to be true, obviously there will be no long term threat to the network to be fixed by a hard fork, otherwise, urgent response should be taken into consideration.

[int03h]

To be clear my words were "CALM DOWN or write some code". There is a difference.

[masonticehurst]

@AliAshrafD I agree, you simply don't write code without a general consensus or reason beforehand.

[QuintLeo]

[Quote] Yes it's official that Bitmain has ASIC now called Antminer E3 with a mere 180MH/s @ 800W. But IMO these are just rebadged version 1.0 of their ASICS that they wanted to get rid of and profit from. Else why on earth they would sell them for mere $800 where a 180MH/s mining rig costs ~$2500 at minimum to build. [\quote] Because they have a very short time available to make their price back before they become nearly worthless when ETH goes proof-of-stake. Also, before AMD cards got into extreme short supply it was entirely possible to build a 6-card RX 470/480 mining rig for more like $1400 that would match the stated performance of the E3 - and at the end of your run with a GPU-based rig you still have all the parts that are usable for SOMETHING ELSE and have a significant resale value.

The recent increase in total network hashrate is EASILY accounted for by GPU sales - it's noteable that when the hashrate leveled off about 2 weeks back, AMD gpus started getting available AND THE PRICE STARTED DROPPING ON THEM pretty quickly. It only takes about 1.5 million gpus per quarter for 1Q and 2Q 2017, perhaps 2 million for 3Q, and a bit over 3 million for 4Q 2017 and 1Q 2018 each. For perspective, AMD and Nvidia combined sold OVER 50 MILLION DISCRETE GPUS in 2017 - and both have been CLAIMING that "miners are the cause of the recent shortage", which on the part of Nvidia is a blatant lie (it's doubtful they sold more than 3 Million or so to miners out of their OVER 30 MILLION TOTAL SALES). AMD has stated that "total GPU sales to miners was less than 10% in 2017" but that GPU total includes on-motherboard chipsets, and probably also includes sales of their various APU and CONSOLE APU sales as well. There is no need for ASIC to have been a significant factor to account for the increase in total network hashrate for the last year (appx) since altcoin prices jumped, and the probability is that Bitmain only has a few "engineering sample" rigs in existence at this time and is just STARTING to ramp up production on them given the MID JULY timeframe before the E3 starts getting delivered.

For some additional perspective - to just MATCH the existing Ethereum network hashrate (leaving out the ethash other coins like ETC) would take over 1.5 MILLION E3 units to be built. Bitmain might have sold that many of their "beats anything else in the industry" S9 miners by now (there's enough room in Bitcoin and SHA256 alts to account for a bit over 2 million S9 units, but you also have to account for sales of the Avalon 721 and 741, internal use at BW.COM of their SHA256 miners, sales of the Ebang units, whatever Bitfury has sold that has gone into mining usage, and "design and use the chips ourselves" on the part of at least one "huge US based mining farm" I'm aware of and possibly others as well that I'm NOT aware of). Keep in mind though that Bitmain has needed almost 2 YEARS to sell that many units, and my actual guess is that they have sold just a little over ONE million units in the last nearly-two-years timeframe. The E3 is NOT going to kill Ethash mining, it's going to take months just to put a noticeable DENT in it - very similar to the situation in Scrypt mining where the older A2/Titan/Alcheminer generation is STILL PROFITABLE TODAY to mine with if you have low-enough electric cost despite being 2-3 YEARS old ASIC technology while the "current generation" ASIC (A4/L3/L21/L3+/A4+/A6) are more than 5 TIMES as efficient. The difference is that the E3 is NOT more efficient than current GPU rigs - it's actually a little LESS efficient than a well-optimized 1070 based 6 card rig (186-189 Mhash common at 700-750 watts draw at the wall) and only a tossup with a well-optimized RX 470/480/570/580 based 6 card rig (164-180 Mhash depending on the specific cards and presuming BIOS modded for improved memory timings AKA strapping, 700-850 watts at the wall ballpark again depending on the specific cards). The ONLY thing the E3 has going for it is twofold - low cost (but with NO residual value) and it's more compact than riser-based rigs (though not a LOT more than some of the 4U "mining cases" can manage, based on the pictures of it to date.

There is NO threat of other companies "easily decoding and replicating" the Bitmain chips - it would be faster and easier to design from scratch. Keep in mind we're talking feature sizes that you need something on the order of a Scanning Tunneling Microscope to even look at - and the field of view on those things is TINY....

BTW - the "usual" size for a Bitmain batch on ANY of their miners has been about 3000 units - there is no way they can get enough chips to manage HALF A MILLION E3 units in a very few months, there isn't enough foundry space left over after Samsung, Nvidia. AMD, Apple, and IBM get through their shares.

Keep in mind that EACH miner Bitmain builds normally has over 100 mining chips in it - it's not like a GPU where the chip is quite a bit bigger but there is normally only one GPU chip per card.

(I don't count memory chips as those have DEDICATED production lines for making them).

[AliAshrafD]

For the record: I'm not speaking of decoding Bitmain's chip (there is no such thing at all) nor I'm afraid of it, instead I'm counting on other companies to check and find out and replicate the tricks Bitmain has employed in E3 (as a whole rig) to keep the costs and the price low. If they fail and let us down, ASIC or any other proprietary technology, it should be slammed back by an immediate hard fork, obvious!

[pilsy]

It's worth it to note that they very well may not be the only one with a working product, https://etherscan.io/address/0x52e44f279f4203dcf680395379e5f9990a69f13c is the address of bw.com ~ who in less than 5 day's has amassed over 2.5% of the global hash power...

[AirSquirrels]

Ignore - I was looking at a litecoin miner.

They do claim to have 50,000 ETH mining rigs available for rent...

[krtschmr]

Email just in

Now you can order even 5 units of the E3!

Dear subscriber,
We have received a very promising response to the Antminer E3 batch. Thank you for that.
Many of our buyers have been unhappy about the limit of one miner per user that we had set to prevent hoarding and to allow more individuals to purchase this limited stock of the Antminer E3. To allow all individuals to buy more units and, at the same time, prevent hoarding, we have now increased the limit to five miners per user.
We have also removed the geographical restrictions to allow our international buyers located in Hong Kong, Macau and Taiwan to buy the Antminer E3. Buyers in Mainland China can still not buy this batch.
Click below to learn more or to order now:
https://goo.gl/QFF2wt
Regards,
The Bitmain team

can we agree, that Bitmain is going furious?

[Arachnid]

Locking at @pipermerriam 's request.

[MicahZoltu]

Closing this issue for housekeeping purposes. People are welcome to continue discussing in this thread, but for additional visibility an EIP should be created or the conversation should be migrated to https://ethereum-magicians.org/