There are rumors and hints coming directly from the IOTA team about some specialized chip based on the ternary logic being developed somewhat secretly. Because Curl hash function uses ternary logic using this chip will make PoW based on the proposed hash function more efficient to run. This question is not about the speculation as to what this ship is. Suppose there is a chip that indeed executes the required PoW much faster with lower energy consumption. I also assume that this chip will only be used for Curl, because even if the ternary logic perhaps has some benefits it would take years to develop compilers and tools to run general purpose software on this chip. Will this specialized chip make IOTA more secure and efficient in the long term? Because IOTA is designed to run on millions on sensors, the chips are going to be very low cost, and the potential attacker will have access to the same chips. Why developing specialized hash function and specialized hardware, when many alt-coins move to ASIC resistant algorithms?
-
1This is purely speculation: I don't believe that the ternary chips would make IOTA more secure (as it is designed to not have a great disadvantage if hashpower of different parties varies). But making IOTA ternary will have the crypto community have a closer look at their hash function, and having a ternary hash function may help later doing crypto processors based on ternary (and making their ternary processor more useful in other areas). And they are more energy-efficient for implementing PoW, so that will give them another use-case.– mihi ♦Commented Dec 11, 2017 at 21:42
-
Continuing the speculation (comment length limit): Therefore, I believe IOTA will help such a ternary processor project more than the ternary processor project helps IOTA.– mihi ♦Commented Dec 11, 2017 at 21:42
2 Answers
Yes, but the increased security is more of a side-effect of the employment of specialized chips. The JINN chips are believed to allow even the most lightweight of IoT devices to become full nodes, because they will contain ASICs that compute the IOTA hash function very quickly and with negligible energy consumption.
The reason this improves the security of IOTA is that such a chip will drastically increase the number of potential IOTA nodes. More nodes means a more complex network topology and higher transaction volume, which contributes to the overall security, health, and utility of the network.
The game-theoretical argument here is that someone who controls a lot of devices with JINN processor (which presumably cannot be used for much else besides verifying IOTA transactions) would not be interested to destroy IOTA, because otherwise they would end up with a lot of useless hardware on their hands.
Essentially, the major players in their field are disincentivised to use their power for nefarious purposed, because that would be self-destructive for them. This is the standard argument why bitcoin is more secure than it appears to be against the 51% attack, and why altcoins such as Sia prefer to go the ASIC way.
many alt-coins move to ASIC resistant algorithms
That is not necessarily a wise choice, at least not in terms of resistance to 51% attacks.
