Bitmain, the dominant ASIC manufacturer for cryptomining, has been on a tear lately, cracking many mining algorithms that had been touted as “ASIC-resistant.” Often the term is confused with being ASIC-proof, which isn’t really possible. With enough time and money a company can create an Application Specific Integrated Circuit (ASIC), and as the name suggests, it’s mostly a matter of designing a chip specifically tailored to mining a particular algorithm. Because CPUs and GPUs are general purpose hardware they theoretically will never be more efficient than hardware that is specifically tailored to a single task.
What is an ASIC?
An “ASIC” is an acronym for “Application Specific Integrated Circuit.” The name pretty much says it all, it’s an integrated circuit (or computer chip) that is tailored to a specific task. In essence, it’s a chip built to do one thing and do it extremely well.
Alternatives to ASICs for cryptomining are Central Processing Units (CPUs), Graphical Processing Units (GPUs), and in some cases Field Programmable Gate Arrays (FPGAs). Most will be familiar with CPUs (Intel, AMD, etc.) and GPUs (Nvidia, AMD, etc.). CPUs are used for general purpose computing and GPUs are used for graphics processing (and lately machine learning). Both can be used to mine cryptocurrencies, with GPUs possessing a clear advantage over CPUs. However, neither CPUs nor GPUs are designed specifically for cryptomining, which is why ASICs are much more efficient.
Why make an ASIC for Cryptomining?
Cryptomining is about securing the network, processing transactions, and of course, making money. Miners provide a service and the more efficiently they can provide that service, the more money they make. The parameters that govern a miner’s profitability go something like this:
- Total Network Hashrate
- Individual Miner Hashrate
- Electricity Consumption
- Electricity Cost
- Block Reward
- Block Fees
- Pool Fees
- Exchange Rate
- Overhead (facilities, labor, insurance, security, etc.)
The total network hashrate rises as more hashing power comes online. This occurs when hardware manufacturers produce more units and either mine with them internally, or sell them to the public. The only time that the hashrate may decrease is when the exchange rate drops significantly and it is no longer profitable to mine, or a different coin is more profitable and miners switch. With an ever increasing hashrate the only way to compensate, and maintain fiat profits, is for the exchange rate to increase, meaning the price of the coin or market capitalization of all coins on a particular algorithm go up. This can also be achieved by creating new coins, but presumably they will have low value and not contribute much to the overall market capitalization. Since we’ve been in a bull market it is easy to assume that mining is always profitable, it’s not.
Risks with ASIC Cryptomining
As with any venture there are inherent risks. Cryptomining with ASICs is no different and in this section we review some of the major risk factors.
The single greatest risk for an individual miner or mining operation is exchange rate risk. If crypto prices fall, mining cashflow drops and the value of capital equipment drops. This can be an extremely painful hit, especially if you need to sell equipment to make up for the cashflow reduction. One way to hedge against this may be to use futures contracts, but that’s a topic for a different article. Conversely, if crypto prices rise, fiat cashflow increases and the value of capital equipment increases—to a point. When prices are on the rise, manufacturers sell more mining equipment into the network increasing the total hashrate and proportionally reducing mining cashflow. You can be sure that manufacturers will sell as much equipment as possible during these boom cycles (or mine with the equipment themselves). In a down-cycle both miners and manufacturers suffer with only the most efficient operations succeeding.
Algorithm Changes (Hard Forking)
In recent months there has been considerable backlash toward Bitmain and ASICs in general. The primary issue is that Bitmain has become more or less a monopoly in the space leading to centralization of hashing power for bitcoin and other cryptocurrencies. In an effort to combat this centralization and perceived abuse of power, developers have proposed forking projects to change the underlying hashing algorithm. This was considered for Sia, but ultimately abandoned. However, Monero followed through with their plan to remove ASICs from their network and forked the project. This rendered the Antminer X3’s useless for mining Monero, but many of the forks and alts still support it. Most recently, Bitmain has cracked the equihash algorithm with the Antminer Z9 Mini. Needless to say, there is plenty of discussion on whether Zcash and others using the equihash algorithm should fork.
All that to say, there is an ongoing risk that when a new ASIC is released that the project may be forked to remove the ASIC from the network rendering the hardware useless, or at best, only able to mine obscure coins or trailing forks (projects that retained the hashing algorithm). If few ASICs were produced and enough market capitalization exists for trailing forks the “defunct” ASICs may still be profitable, but at a lower rate than if all currencies maintained the same hashing algorithm.
Technology advances can render mining hardware obsolete. Historically this had been a major risk, but nowadays advances have slowed considerably and cold hard physics are preventing quantum leaps in performance. Currently there are 7-16 nm miners that are inline with cutting edge performance and using the latest and greatest fabricators. In addition, algorithms that demand high power (i.e. SHA-256/Bitcoin) may not benefit much from reducing feature size. All that to say a giant leap in hardware performance is not expected any time soon. However, one could argue that this isn’t the case for memory intensive algorithms such as equihash, ethash, etc. where power consumption is low and memory is the bottleneck.
This is similar to a hard fork, but usually a consensus change is planned far in advanced and more or less expected. For the most part, this concerns ethereum which is planning a consensus change from Proof-of-Work (PoW) to Proof-of-Stake (PoS). This eliminates the need for mining hardware altogether and instead relies on “stakers” to generate blocks using general purpose servers. In this case the staker must hold the underlying currency as collateral to participate in staking and generate blocks. I won’t go into detail here on the mechanics, but suffice to say that an ASIC miner could become a cashflow producing block producer if they held their mined currency. If they did not, then once the consensus protocol changes they can no longer mine that currency using PoW ASICs.
Of course there are a handful of other risks as well including regulation, natural disasters, fire, and security. Regulatory has already reared its head in China with even Bitmain making plans to relocate to more friendly jurisdictions.
Mining with ASICs is quickly becoming a serious industry, but it isn’t without its risks. It is important to remember that we are in a unique era where payback periods are very short when compared to a traditional business. Even so, things can, and will change. The PoW system is not setup for mining to be highly lucrative forever. Instead it will mature into a self-stabilizing system where low electricity cost and overhead produce a reliable, but small cashflow.