Mining Industry in 2019

Transcript

so I'd like to introduce the second speaker on stage Edward Everson who is head of growth for F T sorry F to Poole place on stage and he will talk about mining aetherium in 2019 hello everyone I can't see any of you cuz the lights are right here but yeah so today I'll be speaking about aetherium mining in 2019 name's Edward Evensen as she said represent f2 pool and during this presentation I hope to kind of reveal some of the opaque

industry that is mining and hope you all understand a little bit more about aetherium mining more specifically some of the contents of this speech will revolve around Prague P of W GPUs and Asics whether all this is still relevant with proof of stake on the horizon and some key takeaways things to consider for a theory of mining in 2019 so let's start off with Prague P o w I'm sure some of you will be familiar with Prague powa already but in order to understand Prague P o w and why it's such a hot topic right now it's really important to look at the rise of aetherium Asics so in July of 2018 bit main releases the

ant miner III and this is one of the big first big Asics for the etherium network in eath hash algorithm and several months later in o silicon releases the eath master 810 in september of 2018 these are about 2 to 3 times more efficient than the regular commodity GPU miners use and so it's not there wasn't too much concern as far as Asics completely dominating the network as well as that they're relatively new over then Lindner another ASIC manufacturer released the announcement of a chip that would be seven times more efficient at the algorithm then these two GPUs excuse me these two Asics and they started to raise some alarm bells started raising questions of aetherium mining and

whether it would start to look like bitcoins Network and mining you know bit main late 2017 early 2018 had basically 75% of the market share and so they started asking themselves do we really want the etherium Network to resemble that of bitcoins in terms of mining so now that we understand that let's go over a little bit of the basics of what Prague powa Chua lis is it stands for programmatic proof of work and basically it's an attempt to optimize GPU mining in a way that destroys the economic incentives to use Asics to mine aetherium this does it in a number of key ways one of the first ways is the

Kukoc hashes are at the beginning and end of the ether algorithm are reduced from F 1600 to F 800 and a word size of 64 to 32 basically what this means is that there is much less instructions that a GPU has to execute when mining the eath hash algorithm and they have native 32-bit data paths so it just makes it much more efficient for them they're also kis 99 calculations added to the film mix stage and while this only takes about five percent of the entire computation for the eath hash algorithm in programmatic proof of work basically you would need to install an accelerator on the ASIC chip which could be a very costly process and it doesn't really add too much efficiency to do so

so that's another reason why the economic incentive of an etherion ASIC would be attacked and then there's many other small details and switches but that basically amount to upping the computational utilization of a GPU to match that of its memory or storage utilization just making GP is that much more efficient so basically if an ASIC were to implement or change its hardware to accommodate prog P o W it would start to very closely resemble a GPU and at that point why invest in the R&D all the extra costs if you could just go buy a commodity GPU and it's almost as efficient so now that we have a short summary of prog pows let's get into GPUs

and Asics in the mining industry go over some pros and cons of each GPU mining you can mine multiple currencies and switch pretty quickly between the two or any currency wanted mine this is been shown with grin recently many GPU miners have switched over to grin because it's pretty profitable for them to do so it can be repurposed for alternate use we used for AI machine learning graphics renderings gaming things like that and because of these various different industries that you can use GPS for GPUs for there's an excellent gray market you can resell them for a lot of value still these standard hardware easy to support reliable people are familiar with them across the globe and they are upgradable

some of the downsides however for the amount of hash they output they consume a considerable amount of power they can't compete with Asics on the same network they're not competitive so this means they basically can't be used to mine some of the most popular crypto currencies to mine such as Bitcoin litecoin D cred for example and for the amount of hash they output they're relatively large so when we think of the amount of space they take up in these large mining warehouses and how like efficient the use of space is it's not that high Asics on the other hand have extremely efficient power usage in terms of hashes per watt output they have a large amount of hash power straight out of the box so in the current iterations

from say--but main what's minor you can have an ASIC minor on the Bitcoin network say that'll have 60 to 70 tera hashes per second which is an order of magnitude above those present on the eath network they're compact for the amount of hash they have so they can fill warehouses quite efficiently and you can quickly recover the cost given a certain electricity meaning that you can invest in a large farm we're a smaller operation and cover the cost currently and about mmm given about a 3% kilowatt hour electricity rate maybe 6 to 12 months however they're very limited in supply speaking with some of these ASIC manufacturers all the orders are sold out really before the announcement

because it's extremely costly to put in these orders at the ASIC chip manufacturers so they need to make sure that they have a sufficient number of customers before they do so so you may have to wait for five months after the announcement date if you're not you don't have one of these personal relationships to get your hands on some many not all s nine being a good example from bit Maine are obsolete upon the release of the next iteration of an ASIC so basically you could have a mining equipment in form of an ASIC and then when the new model comes out you lose like 90 percent of the value of your machine and it's not really that competitive against the new ones so they have a short life span and a low resale value upon that happening there's also a pretty high cost of entry

it takes a lot of startup capital for someone to open up a cig mining operation so high barrier of entry and there's limited coin selection because Asics only really tailored for one algorithm and if you want to get into ASIC mining you know most popular ones Bitcoin D credit like coin and possibly soon etherium one other thing for a con I would add compared to GPUs to this list is that ASIC manufacturers are highly concentrated in China and they're all privately owned whereas GPU distributors and manufacturers say AMD and NVIDIA are publicly traded companies and therefore answerable to shareholders and have an immense amount of pressure to continue

to pump out useful products for the public at large whereas some of these private ASIC manufacturers can simply decide not to sell them Briggs to the public and mine themselves so now that we know a little bit about prog P o w a Sikh versus GPU mining is any of this still relevant with proof of stake so close on the horizon I think we can handle this one fairly quickly well even though you know pure proof of stake Kaspar CBC it's not necessarily coming next month or anytime in the very near future mining is still going to be extremely relevant on the etherium network as far as providing security for that network and there's there's many

reasons why my name is still relevant you know the originally of Casper hybrid the staked coins would be held on the proof-of-work chain the hybrid model will exist for some years before pure proof of stake is implemented my name's not going anywhere and if we care about the security of the etherium network then we still care about mine so things to consider the best case is for prog Pio w maybe the best case is against prog Pio W and some key takeaways from all of this some guest cases for proud Pio w it's Nathan see a theorem is really advocated to be a sakra system you can see this in this yellow paper you can see this on its

website of the etherium foundation from a very early stage they've stated that they don't want to resemble bitcoins Mining Network and with what's happened to it with the introduction of Asics we have the history of the Bitcoin hash wars in recent memory in the summer of 2017 we have the contentious hard work of Bitcoin cache and more recently in fall of last year we have the very very public hash war between Bitcoin cash ABC and Bitcoin cash sv some of you may remember drunken interviews people waving watches around saying they'd be willing to sell them all to protect the network and win the hash war also aetherium network excuse me aetherium

Asics are fairly new to the market the main ones just came out last year so this means that theoretically they haven't had much market penetration yet so sort of forking them out of the algorithm could potentially not pose that much of a threat to the security of the etherium network like it would the bitcoin one few software upgrades are required of mining pools mining farms if we do implement programmatic proof of work so there's not much of a burden on them to change and it combats the centralization brought to mining by the widespread proliferation of Asics a good example of this centralization bit main recently announced that it's going to deploy about 80 million dollars worth of

its mining equipment this season to mine Bitcoin and this kind of puts into perspective the viability for new competitors to enter that space when a monolith that produces these Asics can say that they're going to put nearly you know a hundred billion dollars into mining it themselves and this is another excellent example of how sometimes ASIC manufacturers Pete directly with consumers of their product so some excellent cases for Prague P o w I would also add that should proliferation of Asics reach sort of a critical mass there's also a risk that maybe miners when it comes time to implement proof of stake just continue

to mine the old chain because it's profitable for them and they don't necessarily support the transition so the best case is against it now we don't have the exact sales figures for bit main and in new silicon that sell these Asics for aetherium mining so it does present somewhat of a gamble to remove them from the eath hash algorithm we don't know how much this would decrease the hash specifically and how much more risk it would suppose it would provide to the etherium Network so how much it would widen the attack surface on the etherium network I would also add that programmatic proof of work is a more difficult hashing algorithm and that means less hashes per second and well that's not necessarily the best metric

to gauge network security that coupled with decreasing miner participation could drastically widen the attack surface in a crucial time before this transition to proof of stake in Casper furthermore with Casper on the horizon GPU miners are starting to question the viability of mining aetherium still there isn't going to be as many rewards for them in the future this can be reflected and how many have switched to more profitable ventures like mining be more grin in recent months as we can see from the data this is taken from digital asset research so also there have been a decrease in block rewards several times so with decreasing block rewards more profitable ventures

like mining cream orbit Grinner Beam and taking out Asics from the network all kind of contribute to drastically reducing the hash in conjunction with one another so this could be somewhat dangerous for the etherium Network so what are some of the key figures what are some of the big players saying about all this what are their opinions if you look at Prague POWs carbon vote comm you can see some people that are voicing their opinions and all this there's been about 2.9 million ether staked in favor of proud pow1 note that this is mostly pools and some minors this isn't necessarily everyone weighing in on the issue but it's still pretty shocking

that you know ninety-four percent are dramatically in favor of this some voices against it obviously Linzer who announced the new crazy powerful ASIC they've obviously invested a ton of money into R&D and are trying to sell a product so it's pretty clear why they would protest loudly something like this also some funds have recently come out in some pretty public articles on coin desk and other mediums stating why they think probably pows a bad idea one of which is that Asics aren't the reason for centralization of mining it's really electricity cost I would personally butt back against this because most mining happens in China and they don't have the cheapest electricity in the world it

feels all about the cheapest electricity all mining would be happening in Central Asia Kazakhstan Iran some Latin American countries which isn't the case and also they say that instead of putting so much time and to try to basically making a new hash algorithm we should instead be putting our efforts towards disincentivizing mining pools and large mining farms from becoming attackers of the network and make sure that they are cooperating with developers as opposed to constantly in combat with them also just full transparency you have two people who I represent voted in favor of prog P o W we've listened to our clients and the various miners we work with they

seem to be in support of this and on top of this we think it's important that those miners that aren't necessarily part of this massive network that don't have access to all these Asics also be able to participate in securing the network I've met several at this conference actually who have managed to pull together some resources with some other miners be viable participate in securing the etherium network and they've been involved in the development of etherion for some years now so it's we fear that they get a chance to play a part in securing the network I couldn't go over all the topics as in-depth as I wanted so I encourage anyone that has further questions or that was intrigued

by any one of these topics to come talk to me or any one of us up at the booth or right outside the doors up there thank you for your time and yeah well no matter what happens we'll continue supporting this theorem network whether it's implemented or not Edward so before you go I just wanted to quickly ask you two questions you mentioned something really interesting so I know you you're originally from California but now you live full-time in Beijing so could you comment on where exactly is mining congregated in certain areas of China and if so why yeah so one of the reasons why I mentioned earlier pit main and

then deploying about 80 million dollars worth of hardware the reason why this is happening is that there's something called the wet season in Sichuan which is a Province in southwestern China near the Tibetan Plateau and basically in the spring and summer lots of snow melts off the mountains tons of water comes through there's a lot of hydro plants over there and this means very cheap electricity in the region and during this time at the believe it was Frost & Sullivan their report but about like forty nine point five percent of Bitcoin mining comes from this region during that time and everyone gets really excited for it cheap electricity mining happens all the time and also northern China so northwestern we have gone su and shin heung that are

very popular and it's also just interesting to think about you know we're on the cutting edge of all this technology but we're still so reliant on these natural phenomena like you know we're waiting for water to come down from the mountains to to mine these crypto currencies oh yeah those parts of China and it's very interesting thank you so much thank you okay [Applause]