Decentralization is the essence of public blockchain technology. Keeping mining distributed is a difficult problem. æternity’s team is always on the lookout for possible solutions.

Offering alternatives to existing systems of commerce, governance and finance, blockchain technology has the potential to disrupt industries and provide exciting new opportunities to billions around the world. As the industry grows, the core idea behind the technology — the creation of open, decentralized infrastructure accessible to anyone — is being sidelined and attacked.

One reason for this is the lack of a clear definition of what “decentralization” really means. We published a blog post on the topic that you may find interesting:

https://blog.aeternity.com/building-a-truly-decentralized-blockchain-47ed6bd1f166

Another reason is that creating truly decentralized systems is hard. Following the rules of the Blockchain Trilemma, many projects decide to sacrifice decentralization for speed and security (mostly speed).

Finally, mining poses yet another threat to decentralization. The unceasing evolution of semiconductors makes new, more powerful and efficient chips available on a monthly basis. Cryptocurrency mining is directly affected by developments in the processing units industry. Unwilling to remain dependent on such “externalities”, many modern blockchain platforms are switching to less battle-tested consensus algorithms, such as Proof of Stake. Cutting the ties with the real-world, however, comes with an unique set of challenges. These may prove harder to solve than mining centralization. Only the future will tell.

At æternity, we remain unconvinced that there are any real alternatives to Proof of Work (PoW). We are constantly looking for ways to reduce mining centralization.

Keeping Mining Decentralized

Mining is the engine that powers most public blockchains. It provides security and utility to the network allowing it to create value for its users. If mining becomes centralized, a number of security, economic and even philosophic issues appear. A collusion between parties becomes more probable, making 51% attacks easier to execute. Fewer users receive the financial rewards generated by the network, negatively affecting wealth distribution. Finally, if there are only a few entities that participate the processing of transactions and verification of data, how is that system different than a centralized one (which is much more efficient)?

One development that has threatened mining decentralization is the creation of application-specific integrated circuits or ASICs. These devices can mine much more efficiently than even GPUs and have completely taken over bitcoin mining. Slowly but surely, they are also becoming the primary method of mining for alternative cryptocurrencies. For instance, Innosilicon recently started producing an ASIC for Zcash, while Bitmain’s Antminer for Litecoin has been available for months and its Ether miner was announced in Q2 2018.

Further complicating matters and challenging decentralization is the rapidly-changing hardware world, and the fact that the majority of ASICs on the market are now manufactured by just one major technology company.

Having just one or two central manufacturers of ASICs is a big issue.

Currently, the most viable way of keeping mining more decentralized is still using GPUs. However, GPUs themselves are only produced by two major companies: nvidia and AMD. Nonetheless, GPUs are more universally available than ASICs, thanks to their various applications in gaming, machine learning, complex computations, 3D rendering, etc.

In order to stay on the GPU level for as long as possible (and many thought — indefinitely), ASIC-resistant algorithms were devised. However,

ASICs production is becoming cheaper by the day and GPUs are slowly being superseded.

No mining algorithm is realistically protected from ASICs, yet different algorithms have varying efficiency gaps between GPUs and ASICs. For example, in the case of Ethereum’s Ethash, the gap might only be around 3x, whereas for Zcash’s equihash it is over 20x, and for bitcoin, it is far more still. In all these cases, though, GPU mining is slowly becoming obsolete as ASICs evolve and spread.

At æternity, we too decided to bet on ASIC resistance in order to keep mining on the GPU level for as long as possible.

Cuckoo Cycle was selected as the most promising algorithm for ensuring that ASICs will not be developed in the near future.

However, while they are not available yet, their appearance is much more probable than it was considered a few months ago.

“ASIC” Is Not a “Dirty Word”

Despite negatively affecting decentralization currently, ASICs have a few important advantages over GPUs. They consume significantly less power, produce very high hash rates, and are physically smaller and lighter. In a world where mining is only done by ASICs the biggest issue will be the number of ASIC producers. The development of a competitive ASIC market, however, takes time. What can we do?

Long-Term Solutions to Mining Decentralization

Since ASIC resistance is a short-term solution to mining decentralization, what else can be done? One possible solution is a temporary dual-PoW approach similar to the one planned by Grin.

We believe that a good way forward could be creating open-source ASIC designs, so anyone can build ASICs.

Generally, it is not only the production infrastructure that causes mining centralization — is also the requirement to develop an ASIC designs that can be used for ASIC production. Since Cuckoo Cycle is designed to be the simplest possible PoW algorithm, drafting of ASIC designs could also be less difficult in comparison to other mining solutions.

An open-source ASIC design could enable users to construct ASICs with off-the-shelf components.

This will significantly lower the entry barrier for new ASIC producers, allowing new players to enter the market and start developing ASICs without the need for expensive infrastructure or expensive and lengthy R&D. Ultimately, this could result in the flattening of the playing field for users and companies. Right now, there are about half a dozen projects competing to make the best ASICs for bitcoin. Competition in the ASIC miners sphere will continue to grow over time.

Decentralizing ASIC production is an important step towards keeping Proof-of-Work, with all its time-tested and proven benefits, relevant for blockchain consensus. Decentralizing ASIC production could in turn lead to the decentralization of mining itself.

What other approaches to keep PoW mining decentralized can you think of? We started a discussion in this dedicated Forum topic.

---

Interested in æternity? Get in touch:

GitHub | Forum | Reddit | Telegram | Twitter | Facebook | Mail