Where Are the Ethereum ASICS?


Another eth devs call, another discussion about ProgPoW, a proposed algorithmic change that ostensibly aims to remove asics.

It looks like about 30 minutes was spent on the matter, with discussion mainly surrounding a potential audit, as well as more basic considerations of whether there’s any point to this audit at all, of who is to fund it, of why the Ethereum Foundation (EF) denied a grant for two devs to work on ProgPoW.

“The official statement is just that there are a number of application programs and we selected some. We can’t expect any other statement by the team,” says Paweł Bylica, an eth dev, on why the grant was denied.

Nothing political, some said, but that depends on how you want to read the signal. In any event this closes off with a more fundamental question which is left in the air. Alexey Akhunov, who is working on a complex Eth1x proposal, says:

“I have also been asking whether we would like to keep fighting ASICs or we just do one attempt and we stop at this because I heard ProgPoW has different parameters that we can change.

My question is are we actually going to do this, to keep tuning it to basically do hardforks every six months?

Because this is not defined, the conversations just go all over the place whenever we start discussing it, because it has never been decided.

My concern mostly comes from the appreciation that the bandwidth of changes in the Ethereum protocol is actually limited. That some thing can hold the agenda of this meetings for months. That makes it really hard to do other improvements that we actually do need as well.”

There are many improvements on the pipeline as you probably know. Just as there are different eth devs calls, but this is the main one. It would have been interesting, for example, to have had a brief discussion on what stage the Beacon Chain is at this point. Is a testnet to launch this month as they said? Does it affect Eth1x?

The meeting does move on towards some pretty interesting statements by Akhunov regarding storage fees. Vitalik Buterin, Chief Scientist at the Ethereum Foundation, does chime in briefly on that matter, but when it came to Eth2 research updates, he had left.

Busy guy, with many jobs. A meme figure, a researcher, a leader of sorts, a conference speaker, sort of a spokesman on developments, and a human bound to 24 hours a day.

He was kind enough to find some time for this meeting, but it looks like Hudson Jameson, an EF employee, was more interested in implicitly drilling the attendants regarding this grant rejection by EF.

“I heard that there was a grant application submitted that they got rejected. I don’t know if people know more about it,” he said.

That’s an incredibly appropriate thing to raise in a devs call meeting where time is necessarily limited and there usually is far too much to go through especially at this stage for eth where there is an incredible amount of planned activity.

It is especially appropriate when a number of eth devs have made it quite clear, although implicitly and very politely, that they don’t want ProgPoW. Buterin for example recently said “if attacker using gpu, there’s nothing” in the last discussion about ProgPoW where it was decided the eth dev calls where an inappropriate forum for the matter.

Yet for some reason this keeps being brought up by an individual who has still failed to disclose any conflict of interest, a matter of relevance as considering the proposal appears to be controversial, it is no wonder the Ethereum Foundation doesn’t want to pay for it, but where are these asics anyway?

The ASICS Cryptonomics

No top coin (which isn’t a split coin) has changed its algorithm except for monero, which has as its selling point the fact that it can be mined on CPUs.

They claimed in monero they’ll keep asics at bay by forking every six months, but 85% of its hash is now again controlled by asics and whether they plan to do anything about it is unclear.

Bitcoin has never changed its algorithm and it is largely controlled by asics. A 51% attack of BTC at a practical level appears incredibly unlikely because one would need a massive secret operation to produce the same amount of bitcoin asics as are currently in existence.

Otherwise, someone would have to spend huge sums of money to acquire 51% of current asics and then effectively to burn all that money because the algo can at that point just be changed, making this money down the drain endeavor quite pointless.

From that comparative point of view eth might appear a bit vulnerable because one wouldn’t need to produce the same amount of all gpus in existence nor would they need to acquire 51% of gpus from current miners.

One can instead go to Nvidia or AMD with a gag order and if they do, conceptually, it is difficult to see how eth can respond.

Obviously this is just code and in practice the matter would be far more complex, but technically eth is less secure than bitcoin in a very simplistic comparison because BTC might be less secure if we consider geographical centralization.

Even there, however, the advantage BTC has is a quick deployment of Keccak. So whether it is the Chinese government seizing miners or some other clearly malicious actor, the threat is neutralized in about one or two hours.

It might be a stretch to say perhaps this is why bitcoin is processing about $5 billion a day, while eth is at less than $200 million. This ratio, however, is quite different from that of eth’s market cap being about 20% of BTC while only 5% of value is being processed.

Where Are These Asics Anyway?

The ethereum ecosystem is being told that they’re facing an asics threat and thus they should change the algorithm and urgently, but where are these asics?

Miners vote on ProgPoW, March 2019.

A miners vote on ProgPoW is showing an incredible 100% agreement of those that have voted. It is only 55% of the entire hash that has done so at this stage, but how can it be so unanimous?

If there are any asics, we’d expect at least a 0.01% of a No. That there is none against, suggests either there are no asics, or that they’re not affected by this change, or that the asics miners are completely stupid.

The holders vote is even more incredible because we have 3 million eth voting for ProgPoW and only Gnosis – which is a dapp and has nothing to do with mining – voting against with circa 200,000 eth.

ProgPoW eth holders vote, March 2019.

It’s not easy to explain this at all because it makes no sense if there are asics. A lazy argument might be that asics miners straight on sell, but they’re facing an existential threat. Unless we assume they’re complete wood, then they’d hold off selling eth to vote or they’d buy eth to vote.

There’s another fact. At least 1 million eth that has voted for Yes comes from just one entity as we previously revealed.

That’s a huge sum of money. Why do they care so much about it? One obvious explanation might be they think asics are coming. They want to protect their business, so they push for the change with all they legally can.

That does however reveal a certain fact. Ethereum currently has vast GPU industrial farms with huge resources and they are failing to really contribute by funding projects and so on with their profits.

On the other hand, ethereum does have asics of 2x efficiency or so, but if this data can logically be analyzed then it appears quite clear that such asics are being held by hobbyist miners who are unable to make any difference in any such vote due to their gear producing few eth or hash, compared to 3 million eth in any event.

From there you can easily argue that asics would decentralize the network as it gives the little guy a chance to participate until the big guys build their asics farms. Then once they do, you can have the smart guy compete with them by producing better asics.

And so you have an incredible race that makes the network stupendously secure with a carousel of dominant miners going bankrupt due to inability to keep up in a revolving cycle that deserves not a study paper, but an entire book to decipher.

While with ProgPoW we have a stale industrial farm that is flexing its muscle to push for an algorithmic change that keeps their dominance in an abuse of capitalism with added risks in removing potential defenses to 51% attacks as enjoyed by ETC recently.

That all said, the vote is the vote. It has been gamed, but there are more than 100 million eth which can still vote if it wishes to. If they don’t, then they implicitly agree with the end result.

One can go into philosophical matters of certain entities obviously having more incentive to vote, but ultimately there is no clear cut answer. Nor is there some other way of reaching this sort of decision.

On the other hand, that is sort of saying that one entity can amass 3% of eth’s supply and then gain effectively decision making powers, but that would go into addressing a far more complex matter.

How Do You Fight a Sockpuppet Army?

Censorship is on the rise across the internet in a travesty for the 90s boys and girls with the then dream of an open, decentralized, free space, quickly becoming more and more hierarchical and concentrated into kleptocratic designs that have privatized, and made for business, the open public spaces where we all gather.

Selective enforcement of rules or policies can be seen across many discussion platforms where manufactured propaganda does often run amok with the enforcers usually not doing much about it. Not because they can’t, but because they implicitly or explicitly want to facilitate it.

Few now have not seem themselves banned from some public spaces, and if they haven’t, they haven’t said anything worthy.

An entity with sufficient resources can flood such public spaces and can “buy,” implicitly, explicitly, knowingly or otherwise, the consent or facilitation of an enforcer.

The aim is obviously to push an agenda and the appearance of consensus or public support, in addition to pushing pressure of the style: something has to be done right now.

The only way to fight that sort of “army” is by getting your own army which breaks down the appearance of general agreement if indeed there is no general agreement, which breaks down the propaganda.

Usually, however, the rest are not so invested in the matter and they wouldn’t have much to gain by spending resources on a counter-army, so socks generally get pushback here and there, but the propaganda continues, with the push-back forgotten.

Otherwise said, why on earth are we even talking about ProgPoW when there is no evidence of any asics on the network?

Why is ethereum even having a vote on this matter, when again there is no evidence of any asics concentration?

Why is another eth devs call opened by ProgPoW discussions at the apparent expense of Buterin giving an update on Eth 2.0 research?

At what point was the obvious “no” to ProgPoW given last year, and again in January this year, not sufficient?

Where was the discussion on why asics are not desirable and where was a decision on that matter?

Where was the discussion on just how efficient asics are when this matter was re-raised again after last year’s conclusion was they’re not that efficient at all.

In other words, where is the due process rather than being swayed by what clearly appears to be a sockpuppet army that has now shown its strength by circa 3 million eth and by this matter now dominating eth dev calls for months while it reaches no conclusion nor is it even really discussed at proper depth because as good as no facts whatever have been presented and because obviously a catch-up call limited to 1 hour is not quite the forum for debate.

That all said, the vote is the vote. Some eth devs have clearly been very naive in inviting miners on such matters that concern security or other aspects – even bargaining with them on the new supply reduction proposal – but the vote is the vote.

If eth wants to change its algorithm, and by doing so picks winners and losers without a proper justification, then in many ways it doesn’t matter. Except that there are many things which can be foreseen and many which can’t, so, we won’t quite know whether it does in fact matter until after quite some time.

Giving Craig Wright a podium did not quite matter in BCH either in 2017. This isn’t quite the same, or anywhere close, but it’s an algo change.

There are many very good reasons why bitcoin didn’t go that route. We have seen no discussion of those reasons. How much would the hash fall? If a lot, how likely is a 51% attack? If 30% likely, is the risk worth it? If the algo doesn’t fall much, how much would have this cost in delaying other development aspects?

Who can put a name to auditing this algo? If secret efficiencies are found that concentrate more that 3 million eth, how secure would PoS be? If one self-interested demand is given in now, what would stop other entities from engaging the same playbook?

Who decides this matter anyway? If there is a split, is it worth it to get EF or Buterin to say once more “I am working 100% on ETH,” and on and on.

But the vote is the vote, of one entity. They want this change for unrevealed real reasons that probably amount to something like we have a massive gpu farm and we like it so. Thus the whole network will now have to deal with this irrelevant matter which most probably will make eth a laughing stock in bitcoin twitter at a time when there are some very important, but tough decisions, that actually matter.

But the vote is the vote. We like it or not, one entity or a thousand, gamed or otherwise, it is there open for all. How else do you decide?

Editorial Copyrights Trustnodes.com


Leave a Reply

Notify of