Development of ASIC devices for Equihash made ZCash more centralized, and the idea of switching to a new consensus protocol is already is in the air. We propose the development of a new PoW puzzle that can be efficiently solved by a browser.
This approach has the following major advantages:
Right now, there are different ideas about achieving ASIC-resistance, e.g. memory-hardness and device binding. The current idea is inspired by the recent paper, which argues that modern general-purpose processors are already ASICs for special benchmarks. As far as modern users spend more and more time in a browser, we expect that if an ASIC that efficiently solves our puzzle will appear, it will be in demand by regular users who will want to improve their web surfing experience. Moreover, the current idea may be combined together with other ASIC-resistance ideas, making the new PoW-puzzle even more resilient.
Efficient PoW CAPTCHA
Coinhive PoW captcha was quite popular before the project closed down. Nevertheless, the idea of PoW captcha is still in demand inside a crypto community, which is bored to train the AI of a known corporation be selecting squares with traffic lights. If the disparity between an in-browser and low-level miners will be low, the PoW captcha will provide a non-negligible income for the website owners making it more popular and popularizing the underlying ZCash project.
High decentralized hashrate
Such a frontend-optimized puzzle will allow mining ZCash on a wide majority of devices, resulting in a huge hashrate even if the impact of a single mining device will be small. This will secure the network and bring ZCash back to the original cryptocurrency ideas.
At the current stage, we do not propose the actual implementation of a new PoW in ZCash, but we propose performing scientific research of a topic which should result in:
Apart from the research component, the main expected challenge is the nondeterministic behavior of some functions in different browsers (and their versions). It will require very careful puzzle implementation, especially in its verification part. Another challenging task is to extract and utilize demanded subset of frontend stack in the puzzle, as far as there is no single conventional benchmark there.
The current team contains two independent researchers (Dmitry Meshkov, Vasily Kharin), who have already designed Autolykos PoW consensus protocol and delivered it to production and also have expertise in frontend development. We are also in touch with the frontend developers and designers that may make the resulting product production-ready.
The requested $35k budget consists of the following parts: