How the mining “arms race” is starting to look like private equity in the 80s
By Leo Zhang
Mar 28, 2018
Token-buyers are largely unaware of the multi-billion-dollar effort by mining hardware manufacturers to build faster integrated circuits, but this industry creates tectonic forces that eventually move cryptocurrency markets.
Application-specific mining machines are antithetical to some cryptocurrency project maintainers. Teams from well-known networks like Ethereum and Monero are openly anti-ASIC, with the latter group threatening to fork away from their own network should Monero ASIC machines appear.
This week, rumors emerged that Bitmain, the largest Bitcoin ASIC-maker, is pursuing development of an Ethereum chip which--if true--could balkanize the community (see thread below). They've already announced a Monero miner.
Bitmain's powerful position results in part from the massive and surprising infusion of cash it generated by operating its own machines during the price run-up of 2017. Not unlike the innovative LBO artists of 30+ years ago, it has an unprecedented war chest from a new source--about $4 billion netted last year, according to Fortune.
What does this mean for the industry? Are ASIC manufacturers Barbarians at the Gate who threaten to consolidate the industry? Or will the success of Bitmain draw other silicon manufacturers like Intel or AMD into ASICs, enlarging the pie?
To understand why the question matters, peruse the links below, starting with a thread by ASIC-moderate Ethereum researcher Vlad Zamfir:
Would you support a hard fork that obsceletes ETH ASICs? (Just wondering, this is not a proposal)
"The scheduled network upgrade introduces two major changes. First, a PoW tweak to curb any potential threat of ASICs and preserve ASIC resistance. As a result, miners will have to update their miners (i.e. mining software). Second, the minimum ring size is bumped to 7 (mixin 6). Therefore, pool operators need to ensure that payouts use a minimum ring size of 7 (mixin 6), otherwise they will get rejected by the network."
Some networks like Decred attempt to resolve the question "whether to fork, or not to fork" using on-chain governance systems, which typically rely on voting or ticketing schemes to assign power:
Which network "wins" isn't all about maximum hashpower. It's also about practical use--and one area where bitcoin lags is privacy.
Mimblewimble is a Bitcoin privacy scheme proposed in 2016 by Tom Jedusor. One encapsulating quotation from the paper describes why it was written: "The 'transaction graph' resulting reveals a lot of information and is subjected to analysis by many companies whose business model is to monitor and control the lower classes. This makes it very non-private and even dangerous for people to use."
The Mimblewimble concept has seen some attempts at implementation of late:
Grin team is making great progress & just launched testnet2! New features: * Schnorr-style signatures * Fast chain sync & pruning * Dandelion protocol * Compact range proofs w / bulletproofs * An efficient multi-device GPU miner * Compact block propagationhttps://t.co/Pogdh6OTwB
"But the potential, he emphasized, is huge. That’s because the network, like the brain, doesn’t separate processing and memory. Traditional computers need to shuttle information between different areas that handle the two functions."
"One reason is that Wall Street’s accounting systems are rarely in sync with each other and smart actors have figured out how to exploit these inconsistencies to their advantage. Some of this is nefarious, but much of it is just systemic sloppiness. We will only have a fair financial system when we restore property rights back to the owners of assets in the financial system, and account for them using honest ledgers."