Yesterday I received an unexpected present: a link to a copy of the slides of the presentation that 21inc gave to investors, apparently between October and December 2014, when they were still calling themselves "21E6". (The sender asked to remain anonymous, and I am not sure about the copyright status of the file; so I would rather not repost it here yet. But it seems that several other people, including some of the 21inc competitors, have got a copy too; so anyone who is really interested can probably get it too.) The slides don't have much new factual information, and basically confirm what we already guessed about the 21inc business plans. But they show that we severely underestimated their chutzpah and hype. Here are some random highlights (as far as I can decipher from the slides): + **Mining operations:** They had three relevant mining rig designs in the plans, that would require funding: Codename Qty TH/s kW Cost Deploy Turnoff Profit($) --------------- ---- ---- --- ---- ------------ ----------- ----------- CyrusOne(v2), 7904 2.0 1.3 --- (already on) Apr 2015 ~23,000,000 IO(v1-->v3) 3250 5.2 1.3 2000 Jan 2015 Aug 2016 ~24,000,000 Brownfield(v3) 1900 5.5 1.3 2450 Mar 2015 > Nov 2017 ~20,000,000 > The "TH/s", "Cost", and "kW" columns are per "system", i.e. a mining unit containing many chips. The last column is the expected profit to be made from each set of mining hardware over its expected lifetime. (The slides have some other details that do not seem to be important.) > The first line is the hardware that they were mining with at the time of the presentation; that must be why the "Cost" (as far as investors are concerned) is given as zero. > The second line seems to be an upgrade of their previous mining hardware from v1 chips (which gave 2.7 PH/s total at the time) to v3 chips (which would give 17 PH/s) . > In reality, we have seen that their share of hashpower dwindled through all of 2015, and (AFAIK) they haven't mined a single block in the last six months. Were they still mining with CyrusOne on extra-life, or were they using the upgraded IO which was turned off prematurely? What happened to Brownfield? + **BitSplit chip:** However, their mining operations were secondary; the meat of their plan was the embedded cip, called BitSplit at the time. > The BitSPlit chip (as we suspected) was hard-wired to send 75% of the block reward to the 21inc wallet, whose address was burned in the silicon, and 25% to the user's wallet. The last column, IIUC, is the profit that 21inc expected to make from the 75% cut of the BTC produced by all the chips, over their expected lifetime. > By my calculations, assuming 50 GH/s and no increas in the difficulty, the BitSplit would mine one block in 570 years, on average, and collect less than 2 BTC of reward in that time. So, of course, the chip was hard-wired to mine into a pool run by 21inc, that would spread the user's 25% of those 2 BTC (expected) into a daily regular trickle of a couple thousand satoshis. Their own mining operations would prvied the BTC needed for the pool payouts of all the millions of chips that they expected to be running out there. > They projected to release 3 versions: Model Qty GH/s W Cost Deploy Profit($) --------------- ---------- ---- -- ---- ------------ ------------ USB hub-charger 250,000 38 15 $35 Mar 2015 ~8,000,000 Embedded chip 1,000,000 63 15 $8 Aug 2015 ~103,000,000 BitSplit Inside 10,000,000 20 5 $0 Oct 2015 ~292,000,000 > The "Qty" is the expected number of units sold. > In the above "USB hub-charger" model was a USB charging unit, roughly 3 x 2 x 1 inches, with 2 USB outputs and a mining chip inside, produced by 21inc themselves "to seed the market". > The second line, which I called "Embedded chip", seems to refer to discrete BitSplit chips provided by 21inc and included in consumer devices (like routers etc.) by OEM manufacturers. > The "BitSplit Inside" model would be the BitSplit integrated into the chipsets of other manyfacturers, and manufactured by them. Its cost is listed as "$0" (for 21inc) because they expected those manufacturers to shoulder the cost of manufacturing and integrating the mining chip. > Apparently the market-seeding "USB hub-charger" was later replaced by the "Bitcoin Computer" (aka the PiTato). In one slide it is called "multifuncional BitSplit device", and depicted as a sleek shiny black box, the size of a cigarette pack, with a power cable and 2-3 USB or similar outputs. If that is supposed to be the PiTato, presumably they had not yet realized that a 15 w computer would need a cooling fan with a miniature [wind tunnel](https://en.wikipedia.org/wiki/Unitary_Plan_Wind_Tunnel_%28Mountain_View,_California%29) on top. > In the last two entries, the manufacturers (not the device owners!) would be rewarded with the 25% slice of the BTC mined by those embedded chips. As an example, the slides say that a manufacturer who produced one quarter of the embedded BitSplits would get the 25% cut on the BTC yield of those chips, that was estimated to be between 2 and 4 million dollars per year of revenue in 2015--2018. Those numbers are based on the following predicted mean BTC prices: $350 for 2015, $1000 for 2016, $2200 for 2017, and $5500 for 2018. > So their main business plan was fantastic: the OEM and chipset makers would pay the costs of producing and integrating the chips, the consumers woudl pay the cost of operating them, and 21inc would get 75% of all BTC mined by them, expected to be worth 400 million dollars. > It makes sense to invest 100 million in that plan, right?