CPU to ASIC to Clouds: Crypto-miners still changing Pelotons

Processing muscle matters a lot when you need really-brute compute force for mining something as mathematics-y as crypto-currencies. Mining hardware has come a long way in a short time, from CPUs to GPUs to dedicated ASICs and now Cloud-mines. But are we sure we are not just jumping hoops?

Despite the hills, the sly turns and the steep parts; a good peloton always manages to shine – because every cycle is doing something there. It works to fight the drafting. It brings in an aerodynamic advantage. It keeps the rivals away. Miners would know what these cyclists go through. They climb these inhospitable ascents too. Mining a block takes sweat, patience and practice. But it also takes a good cycle.

Cycling in the crypto-alps was fun but it took time and heat. CPUs were the cycles that everyone rode. No one athlete had a better gear or a shinier handle. Then, some of these cycles slowly made way for sporty bikes. The young ones embraced GPUs that leaned in well to the speed, ease and accessibility that individual miners needed to get better hash rates. But these hash rates could get even better. Enter ASICs. With Application-Specific Integrated Circuits, cut and stitched specially for crypto-mining, the game changed again. ASICs seemed to have cracked, even if for an enviable while, the need for processing complex crypto-algorithms while juggling the efficiency, power and usability metrics to some level.

The catch, however, lurked somewhere. The upshot of scaling up processing stamina in the form of ASICs was that now individual miners could not really afford these top-crust machines.

Until GPUs reigned, the race still hinged upon the contribution of every cyclist in the peloton. ASICs were these sturdier and luxury-racing bikes that now only a few cyclists could afford. What’s wrong with that? The hills were still the same. The terrain was still as stubborn. Except that – now the huddle was breaking up. Into the have-s and the have-not-s.

Nike vs. Jute Shoes

Decentralisation – the edifice of blockchain technology – was the very part that was being sacrificed for other goals like speed, processing endurance, efficiency and hash rate. These goals were not trivial but to chase them at the cost of decentralisation – now that ruffled a lot of feathers. And rightly so. It ditched a lot of small cyclists. And wrongly so.

No wonder, players like Monero kept working even harder to still make GPUs relevant for individual miners. There are also some mining pools such as Ethermine, SparkPool and Nanopool that mainly serve graphics card mining machines, primarily used for mining ETH and other digital currencies. But the wallet-size and equipment-factories that big mining pools could ‘swag’ was still out of question for small miners. The pelotons were now changing into pools.

There was more than money into play here. Electricity – a key determinant of how well a mining set-up sustains in terms of carbon-maths and profits – was, inevitably, also thrown into the mix. So now specific geographies and business-armies started imperial-ising the mining playground. One had to either have cheap electricity or the moolah to afford lots of processing-Porches (ASICs) – and Voila! The mining lead-cyclist emerges.

But for how long before he skids? The draft is flowing in a new way in the last few months. Things are changing.

Tour De Mining

As per some news reports, a few current mining-pool Walmarts could get wobbly due to regulatory and energy-efficiency storms. It has been heard that inspections have begun in China’s Inner Mongolia autonomous region for wiping away ‘illegal’ bitcoin mining operations soon. It is also being interpreted as a strong example of the nationwide phase-out plan on the bitcoin mining that could shake up a lot of data centres that work for bitcoin miners, especially those that have been enjoying preferential electricity prices and tax breaks.

Reports also flitted about that some Iranian authorities had, this June, seized about 1,000 Bitcoin mining hardware from two abandoned factories and the government is seriously considering the ripples of electricity subsidies and mining-power consumption.

If mining pools are facing crackdowns, mining manufacturers are having their own struggles too. This June Ebang International Holdings Inc., one of the world’s top three mining hardware manufacturers failed to list on HKEX the second time. Meanwhile, the overall scenario of mining difficulty as well as mining hardware is also exuding patterns worth noting.

As per a TokenInsight 2019-Q2 Cryptocurrency Mining report, we can expect mining hardware with higher computing power, launched by mining hardware manufacturers, to be put into operation in the remaining half of the year. Going by the market recovery seen in the first half of the year, mining equipment purchased by miners in the first and second quarters of 2019 will be delivered successively in the third and fourth quarters. All this could easily galvanise into a jump in the computing power in the second half of 2019, increasing the overall difficulty of mining. But with hardware shortage as a huge stone that keeps rolling.

Currently, the Bitcoin mining difficulty on the overall network clocks in at around 9.98 T and the average computing power is 74.35 EH/s.

If we correspond this to compute power and hardware costs, the report shows up interesting translations. Assuming that the price of Bitcoin will move up steadily, the minimum average increase in computing power seen from the difficulty adjustment cycle can touch 5 per cent in the third and fourth quarters of 2019.

So whether we go by conservative predictions (mining difficulty will increase to 14.74 T, and Bitcoin computing power will increase by 48 per cent at the maximum, and reach to 109 EH/s) or optimistic predictions (mining difficulty will increase to 17.14 T, and Bitcoin computing power will increase by 72 per cent at the maximum, and reach to 127 EH/s) – there is no disputing the trend that mining difficulty is taking on.

Interestingly, this could start to redefine the mining hardware space in small but sublime ways.

It has been seen how High Power Efficiency Bitcoin Mining Hardware continued to perform well in the second quarter of 2019. But along with that, a dramatic surge in the prices of popular mining hardware was also witnessed during the second quarter. To add to that, the demand for mining hardware exceeded total supply. Mining costs increased and this further shortened miners’ recovery cycles the second quarter.

Yes, the industry is changing the winds. Major mining hardware manufacturers were vying strongly to launch their own powerful SHA256 mining machines. But popular mainstream mining hardware like Antminer S17, Ebang Ebit E11+ and Innosilicon’s T3 43T were seen out of stock during the second quarter. As that happened, there was something else bowing out from the market – mining hardware with low and medium computing power. A lot of it was withdrawn from the market, citing product iteration or unshelving as reasons.

New cycles keep entering, nonetheless. So do new inclines.

Push the Pedal. Smell the Dragon Mint

Recently it was seen that a really large minefield is expected to come up in Russia (slated to have a total investment of about USD 7.3 million, and spread over an area of 4,000 m2) that will house 3,000 mining hardware and will have an energy capacity of about 20 MW. Big players like Bitmain are rolling out new products like a Baker on a spree. Antminer T17 boasts of 73.9 per cent more computing power and 17.9 per cent decreased energy efficiency ratio against its previous siblings. And of course, challengers to Antminer are not giving up their dragon-fires on better energy and efficiency any time soon.

That said, the price of mining hardware has headed only northwards and several times in the second quarter as seen with dramatic increases in the price of Bitcoin. The TokenInsight report highlighted a key point that miners’ overall revenue and the recovery cycle do not simply form a linear relationship.

It is important to also take into account other numbers too – like the price per unit computing power, power efficiency, mining difficulty etc. That should explain why even when the prices of most mining hardware have nearly doubled, the overall recovery cycles of mining hardware only shrunk in the second quarter with the rise in the price of Bitcoin.

Turns out that mining hardware with short recovery cycles generally have higher energy efficiency ratio per unit of computing power

Experts note that when miners invest in mining hardware, it is prudent to also think of the cost per unit of computing power and power efficiency ratio per unit computing power. Electricity price (whether as a variable cost that affects the lowest operating costs or as a change that impacts the recovery cycle of mining hardware) does change the calculator-numbers for mining revenue.

That is, perhaps, why these advantages fare better for the mining pool in gaining more computing power and having a Matthew effect. But that also means that the mining pool outside the top 15 may be gradually eliminated from the competition.
Time for Echelons then?

Let’s Change The Formation, Says Cloud

With all these shifts changing the way hardware supply, hardware prices, hardware recovery, mining operations and profitability work – could cloud mining be the next one to occupy the musical chair that ASIC usurped from GPU?

Hosting hardware and allowing individuals to rent a part of the action instead of buying hardware themselves – looks good on paper, to start with. Now every cyclist, even a small one, is given attention.

Ask TokenInsight Analyst Daniel Qin and he avers that cloud mining could be regarded as the entrance of mining for individual investors as they do not need to invest heavily in purchasing mining hardware and operating mining facilities. “Therefore, with an increasing understanding of the mining market, cloud mining tends to be more acceptable for investors. Plus, there may be more forms of cloud mining emerging in this market.”

But, like every other alternative, this one would also struggle with juggling too many balls in the air – mining profitably but without spewing out too much carbon, while keeping scam-wolves at bay and with ensuring that efficiency is kept on a firm grip. What if this time cloud is more of a financial scheme than a technology model (that brings economies of scale to IT industry customers)?

As Qin seconds, cloud mining has a high possibility of evolving into a financial product. “It really depends on the mining difficulty, Bitcoin price as well as other variables that determine the profitability of mining hardware. However, there is one thing promising that mining hardware is going to be more energy-efficient and environmental-friendly. Also in the future, there can be higher hashrate hardware appearing in the market.”

Interestingly, this July itself saw Cloud cryptocurrency mining service HashFlare, that begun in 2013, declaring a shut-down of mining services and hardware on its existing SHA-256 contracts. Reasons surmised – difficulty in generating revenue.

So even cloud won’t be spared from the tough terrains that other mining formations have confronted till now. A provider can configure the hardware, maintain uptime and make it easy for individual miners to earn from mining. But won’t it have to select the most efficient and reliable pool somewhere? What about energy-footprint and cooling challenges/opportunities? This form of mining can certainly restore the balance for individual miners. But would democracy equal to decentralization?

Every cyclist is given attention. But is s/he given a role too?

Peloton or Echelon – the race is pointless as long as the industry chases to find the balance among decentralization, processing power, efficiency, carbon responsibility and profits.

Still spinning on our wheels, until then.