This is the latest Bitcoin miner from Spondoolies, the SP50 Pickaxe.  This is released only for large commercial miners according to Spondoolies.   Looking at the whooping 16Kw needed to power this beast, they are correct.  You will need some serious power distribution and cooling to even handle a single unit.  The hashing power is the most impressive aspect, tipping the scales at 110 TH/s of mining capability.  They state it only used 0.15 / Watt per Gh/s, this makes it the most effiicent mining solution available to the commercial mining market.   We assume is still uses that clean UX admin interface and the design is pretty clean, getting its inspiration from the current Blade server chassises from Cisco.   The blades look to be the mining modules, 10 blades in all.  That means they have gotten 11 TH/s in each blade, that is a major improvement over the SP35 Yukon. 

Specifications from their website:

Item	SP50 "Pickaxe"
Effective Hash Rate	110 TH/s ± 10%
Nominal Power Consumption	16.57 KW
Power Efficiency	0.15W/GH/s
ASIC	800 Spondoolies-Tech Pickaxe ASICs
Controller	TI Sitara 1GHz (based on Beagle Bone Black)
Controller Board OS	Linux (embedded)
Mining Software	cgminer with custom plugin
Form Factor	11 U rack mountable (mounting ears provided)
Network	Single 10/100 Ethernet port
Fans	6 X 120 mm
Power Supply	10 x 16 KW
Input Rating	180 - 264 VAC
Ambient Operating Temperature	0 °C to 35 °C
Emissions Compliance	FCC / CE
Safety Certification	CE
CFM	1350

 

Bitcoin mining in 2015 is now serious business.   In the early days of Bitcoin mining, that we participated, mining was a completely different beast when is comes to income earned compared to expenses to mine.  A majority of the mining capacity was multi-GPU mining units with just a sprinkling of pre-ASIC hardware that still carried a large premium.   Much has changed since those days.  We are now in the 4th Generation of ASIC Bitcoin mining hardware and over 400 petahashes of capacity. 
 
The modern Bitcoin miner is of a much more professional nature while they understand you not only need efficient Bitcoin mining hardware but a comprehensive hosting plan to maintain their mining capacity over the duration of their investment.  At the heart of these calculations is the power efficiency of the Bitcoin miner.  What this boils down to is the how many watts does the miner use for each Gigahash of processing power.  Take the newest offering from Antminer, the S7, it takes only 1,210 watts  to produce 4.8 Terahashes of mining power.   This equates to 0.25 Watts of power per Gh/s.  That is a huge leap from the first generation of ASICs like the Avalon that was a 8.3 Watts per Gh/s.  The newest generation is 32 times more efficient than when ASICs first started mining Bitcoins on the Blockchain.  This trend will continue. 
 
The reason this is so important is because as the difficulty rises, the more inefficient miners will have higher expenses to generate their Bitcoins.  With rising difficulty, you can imagine miners being trapped on an island with rising water (difficulty), the more efficient your mining hardware, the higher land your operating is on.  Once your underwater, this means that your hardware is costing more in hosting expense than the underlying Bitcoins it generated in a given set of time.  This means that if you are mining in this manner, you are speculating that the price will increase long-term and that is worth mining at a loss in the present.  This situation you will find become increasingly rare.  Many large miners we speak with are quite aware of this and the mining operation needs to stay in the black.
 
The other force pressing on this calculation is the hosting expense of running tens or hundreds of Bitcoin miners.  Along with this march of increased mining efficiency, it has along increased the amount of heat and power required to effectively run these miners.  This is where traditional data-centers and Bitcoin colocation startups come into play.  The power density needed to run a medium or large farm outstrips most commercial buildings and will require extensive and expensive upgrades to make capacity available to operate the miners.     Their most important expense is what the utility charges them for power.   Power rates vary all across the world and even state-to-state in the U.S and Europe.  Hosting companies need to charge a price to Miners that covers their utility costs, along with other operational expenses.
 
That means that because of the dependency of miners on using some form of hosting, means there is a price floor above the raw cost of electricity that equates into the mining profitability calculations that determine the overall performance of the investment.  Currently in my calculations, a price under $210.00 USD per Bitcoin is so low that long-term mining is not profitable with the current generation of miners.  

Mining is a zero-sum game, you get Bitcoins via mining by reducing the amount of Bitcoins going to all the other miners that were mining before the newest miner came online.  This cold war arms race will continue until the Bitcoin mining in its current form is consolidated into a few large mining interests that are likely privately funded with small independent miners becoming extinct over time.  Once we make the move to 14nm-16nm ASIC chips for Bitcoin mining, I see a real possibility that no more Bitcoin ASIC miners will be offered at retail to anyone wanting to get into mining.  The engineering expenses alone will make the risk pretty high, so anyone backing that will want certainty and that will mean not relying on the general public to recoup the expense.   Stay tuned, we will be writing more on this subject and the next 3 years will be very defining for the future of Bitcoin mining.