The cryptocurrency mining ecosystem is about to become very exciting in the coming months, so we think it merits a Q4 2018 bitcoin mining projection. The last time we wrote a projection blog, we theorized that for the most approximation of the global hash rate – you need only look at Bitmain’s production schedule. The update comes as we move to a world where that may no longer hold true, and thus a new model is needed.
As Bitmain announces the next-generation bitcoin miner to replace the flagship, paradigm-defining Antminer S9, we once again enter a new paradigm. This time, however, the landscape looks a bit different as a new slate of competitors threaten Bitmain’s altcoin ASIC monopolies.
Before we get into what’s in store for the remainder of Q4 2018 bitcoin mining and the beginnings of 2019, let’s look at how we got where we are today.
Bitcoin Mining Q1-Q3 2018
Our Batting Average – How close were our last projections?
We published our projection on January 4th, 2018 and included a series of projections based on the disclosure of Bitmain’s production schedules by TSCM. The image below was our estimation of 500,000 Antminer S9s being produced and hooked up per month run through our model. Let’s see how we did:
We said there’d be 500,000 S9s added to the hash rate each month which would result in a daily reward per S9 of roughly .0004BTC per day. At the time of writing, the actual daily reward for an Antminer S9 is just under .0005 or .000496. We did in fact hedge and say it might be a bit lower if you thought Bitmain would produce more altcoin miners from those numbers, or if the actual failure rate of the chips was higher than advertised (which would have likely been exaggerated by TSCM). Not only did they choose to flood the litecoin and dash networks at this time with Antminer L3+s (and later overclocked versions dubbed the Antminer L3++) and the Antminer D3, they also came out with several other miners including their obelisk.tech-killing siacoin miner the Antminer A3, their extremely contentious monero miner the Antminer X3, as well as miners for ethhash and bytom.
The hash rate grew from 13.7 EH/s to about 53 EH/s at the time of writing. That addition of 39.3 EH/s, if all or mostly S9s which we strongly believe they are, implies that a little over 2.8 million S9s where turned online in three quarters from Q1-Q3 2018, or roughly 312,000 a month. A little off, but we were in the ballpark at the same level of magnitude.
So for those playing at home, add this to the list which already includes the flooding of the X11 (dash) hash rate, and the punctuated ruination of Obelisk.Tech upon the release of the Antminer A3, of things we’ve called and been generally right on.
What else happened over the last 9 months?
Miners don’t pay attention to the price, so nobody noticed when the price went down.
While that’s a well-known lie some miners like to tell themselves, it apparently proved true as the hashrate continued to explode as if the price remained unchanged from its January levels. Remember, we just said 2.8 million Antminer S9s were plugged in over the last 9 months. The hash rate had its first difficulty decline in years on July 20th, 2018 as the first indication that the growth rate might be growing and the current paradigm (mining with an S9) might be coming to an end when coupled with the price decline.
It was over this time period that Bitmain continued to weave its tentacles through other algorithm ecosystems; they expanded their grip on Scrypt (litecoin) and X11 (dash) to the point where it became undesirable for many to continue mining. In fact, in the recent months we’ve seen a pronounced decline in hash rates for these two algorithms:
In addition to the consolidation of those two algorithms, Bitmain spread out to several new ones including a complete capture of the blake2b algorithm (Siacoin), Cryptonight (formerly Monero), bytom, and Ethhash algorithm ecosystems. Bitmain’s strategy obviously seems to be driven by a few principles:
- Leveraging their economies of scale to produce units for new markets (algorithms)
- Continuously expand tactically into spaces where competitors are attempting to get a foothold in the market (e.g. Blake2b/Siacoin, Blale256/Dash)
- Prove who really controls blockchains – the miners not the devs (e.g. Cryptonight/Monero episode)
Upon reflection, we realized we didn’t cover the drama that was the release of the Antminer X3 miners, whose original purpose was to miner Monero but the developers decided to fork the algorithm in an attempt to brick the network for the first time (note: siacoin devs are considering the same to provide a remedy for their delivery woes for early Obelisk SC1 purchasers – a whole ‘nother story in itself). It is quite interesting that the fork resulted in four separate projects, each of which hold some value (a result of it being off the radar enough to be in a quantum super-position of a legitimate and illegitimate shitcoin) once again exposing the most overlooked question in blockchain: who holds the power – the miners or the devs? Perhaps we’ll circle back to that story at a later date.
Bitmain is now threatened on two fronts – new entrants seek to release more efficient SHA-256 bitcoin miners which threaten its core revenue stream, as well as new-paradigm altcoin miners which threaten its wider grip on the ecosystem which to date has allowed the prevention and immediate incapacitation of any competition. Perhaps this was the ultimate reason for the push to IPO and cash out – the writing may be on the wall.
Looking Forward to 4Q 2018 Bitcoin Mining
The next quarter – October through December of 2018 – will prove to be very interesting. A few new entrants to the mining space are set to release new miners for bitcoin, as well as altcoins, which are purported to be an order of magnitude faster than the current fleet of Bitmain models. Let’s look at each algorithm to understand the landscape:
SHA-256 – Bitcoin / Bitcoin Cash
It seems as if everyone is producing a SHA-256 machine slated to be released sometime in Q4. Some companies have been a bit more transparent than others, but all are looking to the Obelisk.Tech fiasco when considering whether to be vocal about their hardware development plans. If you’re thinking about mining bitcoin (or bitcoin cash) in the coming months, here is our ranked list of hardware you should consider:
- Bitmain 7nm Miner – 42W/TH – Price not Disclosed – Release date not announced
- Pangolin Whatsminer M10 – 65 W/TH – $1788 – October 15th, 2018
- Innosilicon T2 Turbo+ – 66.6 W/TH – $1568 – Late 2018, Early 2019
- Ebang E10 – 90 W/TH – $4400
- GMO B2 Miner – 81.25 W/TH – $1999 – Late 2018 (likely vaporware)
- Fulijuntuan Titan 2 – 29W/TH – $2500 – Late 2018 (definitely vaporware. Just look at it!)
If Bitmain’s being honest with their next generation capabilities, then it appears Bitmain is set to retain control of the SHA-256 algorithm market. Cost and release date will be two factors which could threaten this monopoly, but it appears as if Bitmain is able to generate a 2.3x improvement over the current model which should be sufficient to eventually control much of that hash rate.
Important to note are the new entrants:
- Whatsminer, founded by a former director of Bitmain’s hardware engineering division
- Innosilicon, an incumbent in the hardware manufacturing space
- GMO, a giant Japanese telecom somehow foraying into the space
- The usual cast of “not-Bitmain” resellers of overclocked equipment attempting to mislead consumers.
Whatsminer and Innosilicon are both set to release miners that are more efficient than the ~96 W/TH Antminer S9 which means the Q4 2018 bitcoin mining hardware distribution is going to change a bit. As we don’t believe Whatsminer or Innosilicon have the production capacities that Bitmain has amassed, its unlikely to render the Antminer S9 unprofitable until later in the quarter, if at all. These units will then become obsolete once a true 7nm miner is released.
Scrypt – Litecoin
Pronounced “s-crypt”, the algorithm behind litecoin’s market for mining hardware has been dominated by Bitmain since the release of the Antminer L3+ in May 2017. As mentioned earlier, this is one of the hash rates that’s been flooded with equipment to the extent that the hash rate has actually declined appreciably in recent months. As such, you don’t see many mining companies rushing to develop Scrypt miners.
That being said, FutureBit is releasing a more unconventional, low power miner which is supposed to have slightly higher efficiency than the L3+ (and L3++ since it’s just an overclocked L3+). The FutureBit Apollo – if enough are produced and put into production – could begin to wrest power away from Bitmain in this ecosystem.
- FutureBit Apollo – 1.42 W/MH – $274 – Shipping October 2018
- Antminer L3+ – 1.58 – $193 – Available for 1.5 years.
X11 – Dash
It does’t seem like anyone is really focused on bringing a new dash miner to the market. The Antminer D3 appears to be the remaining champion due to the aggressive flooding of hash power to the network in late 2017.
- Antminer D3 – 63 W/GH – $133 – Available today
Equihash – Zcash
There is an active battle unfolding in the Zcash market, which might be why the Zcash price isn’t being hammered as hard as some other altcoins. The Innosilicon A9 is currently the most efficient zcash miner on the market (and we should know – we host a lot of them). The Innosilicon A9 marks the first time a competitor has beat Bitmain to market with a more efficient unit, as the Antminer Z9s and Z9 minis have only just begun to ship and are actually slightly less efficient than their Innosilicon counterparts, though at a higher upfront cost. This unit may give Innosilicon’s investors the win they need to stay in the game against Bitmain.
- Innosilicon A9 – 12.4 W/KSol – $3629 – Available today.
- Antminer Z9 – 23 W/KSol – $3091 – Available today.
- Antminer Z9 mini – 26.6 W/KSol – $793 – Available today.
Blake2b – Siacoin
While Obelisk.Tech was the first to announce their intention to bring a Blake2b (siacoin) ASIC to market, they were ultimately beat by the Antminer A3 and subsequently the Innosilicon S11. Both of these units perform orders of magnitude better than even the Obelisk SC1 was purported to, rendering the units obsolete prior to delivery. While the developers muse over the idea of forking the chain to provide sanctuary for people who pre-ordered units from them there is an active mining community around the algorithm. If what happened to Cryptonight is any indication, regardless of whether there is a fork – it will continue to be.
The Innosilicon S11 once again outperforms the Bitmain Antminer A3, making it the miner of choice for siacoin in Q4 2018. In a few months, it will be possible for Innosilicon to have wrested control of Equihash, Cryptonight, and Blake2b away from Bitmain, with the battle for Blake256 (decred) looming on the horizon.
- Innosilicon S11 – 321 W/TH – $1150 – Available today
- Antminer A3 – 1,564 W/TH – $ – Available today
- Obelisk SC1 – 909 W/TH – $1999 – Unlikely to ever be delivered without threatening a lawsuit
Cryptonight – Monero forks
Bitmain first released the Antminer X3 to the surprise of the mining community as an attempt to produce an ASIC for an “ASIC-resistant” blockchain. The developers stuck to their principles and changed the hash function used to mine Monero, which should have bricked all of the Antminer X3s. Instead, four new projects emerged that retained the chain history and stuck to Cryptonight as the mining algorithm of choice. As such, a battle for Cryptonight supremacy has ensued between Bitmain and Innosilicon.
Both miners are roughly the same efficiency, with Innosilicon eeking out slightly better specs at a higher price point.
- Innosilicon A8+ – 1.93 W/KH – $1499 – Available Today
- Antminer X3 -2.11 W/KH – $1155 – Available Today
Blake256 – Decred
The Blake256 mining algorithm used for mining decred is about to become the proving ground for mining hardware manufacturer supremacy. Bitmain, Innosilicon, and Whatsminer all have plans to release a decred miner in Q4 2018. The Innosilicon D9+ is available today (unlike the Obelisk DCR1) but is soon to be supplanted by the recently announced Antminer DR3. All three of these miners will be put to shame by the significantly more efficient Whatsminer D10 if and when it launches next month.
- Whatsminer D10 – 50 W/TH – $4198 – Releasing October 25th, 2018
- Antminer DR3 – 180 W/TH – $639 – Releasing October 2018
- Innosilicon D9+ – 439 W/TH – $1499 – Available Today
- Obelisk DCR1 – 416 W/TH – $1999 –
May or may not be released due to legal woesedit: apparently people are in fact receiving their DCR1s within September 2018.
Q4 2018 Bitcoin Mining Locations
We hope this list has provided a good snapshot of the mining marketplace and ecosystem at the start of the fourth quarter. One of the obvious questions this analysis begs is – where will all of these units be hosted? IT’s unlikely that miners using previous-paradigm equipment (e.g. the Antminer S9) will decide to unplug their machines the moment a better unit hits the market. In fact, we know this to not be true given the number of individuals still mining with Antminer S7s in 2018. Some people are true believers and really don’t consider the daily price.
That being said, there will continue to be a huge demand for mining colocation throughout the remainder of the year. If you find a space or provider you’re happy with – make sure you lock it in as prices for hosting may surge as demand increases. There are a number of large farms coming online in the coming months – from the few we know of coming in to the US South, expansion of the footprint in upstate NY, government-sanctioned mines in Quebec and Northern Canada, and the few new mines opening in Siberia, Georgia, and northern Eurasia.
If you do decide to purchase any of the miners above, we hope you’ll consider the Miners’ Union for your colocation needs. Until next quarter, happy mining!
 We were very careful not to use the word “profitable” here. Often people discuss the profitability of mining as the ability to cover your costs in USD from cryptocurrency proceeds. This isn’t a fair definition, as it almost always involves some crypto-to-fiat gateway in order to pay utility bills which have various costs depending on who you are and what exchange you rely on. Presumably the energy companies are not progressive enough to accept crypto directly quite yet. Another way to consider profitability is to consider the future value or discounted future value of the cryptocurrency you’re mining, which relies primarily on your own assessments. All we know is, some people either decided mining was no longer for them, or the machines themselves went on strike (fire, flood, malfunction, or a robot strike to protest working conditions).
 This is one of the difficulties that arise when the developers are also hardware manufacturers. Perhaps a blockchain becomes too “centralized” when one party controls more than one leg of the users/developers/miners tripod.