recycling the SHA-256 ASIC rigs

HadrienHadrien MaltaMember Posts: 17 ✭✭
Hello world.

We can observe an geometric growth in the computation power dedicated to the processing of SHA-256 with ASIC.
All this goes to Bitcoin and some of it's memes.
But can it hold?
No ecosystem have such a growth rate without failure in time.
Should we anticipate, and plan ahead another use for those ASIC rigs?

Can such computation power be used for Ethereum?
I am far from the level of Ethereum coders, but I am confident SHA-256 can be used in Ethereum in a way that authorize the recycling of the old (and future old) ASIC rigs.

Same logic applies to "script-ASIC-mining"

I understood that, by design, Ethereum forbids the use of ASIC.
But I believe as well, after reading the papers of Vitalik, that Ethereum will take over Bitcoin (and by extension over the other alt-coins).
I understood that the integration of Open Transaction is as well to permit having Contracts including any cryptocurrency.

My point is how to recycle what will soon or later be a HUGE obsolete distributed supercomputer dedicated to SHA-256.

Thank you for your time.


  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    AFAIK essentially what those mining ASICs do is useless for anything other than mining.. Could think about trying to include it in mining, as a secondary aspect of it, especially if you can figure out a way they they essentially cannot get any power of the network regardless of the magnitude of the mining..

    However, it seems to me like any sort of attacker will have no problem getting at some ASICs, i dont think there is really much use other than making those machines useless. Not really our responsibility, and I reckon doing so contributes to the value of investing in those ASICs and thus centralization risk of bitcoin. So i dont think it is a good idea.
  • chris613chris613 Member Posts: 93 ✭✭
    @Hadrien‌ I too lament the colossal waste associated with mining rigs that serve no other purpose than performing a double-SHA256. Re-using them for other purposes is tricky, though, and brings all the original problems along for the ride.

    Part of Ehtereum's mining algo design (which is not yet finalized) is to attempt to use programmatic logic as part of the problem to solve (instead of just one hash function) so that any "specialized" mining hardware would actually be a powerful general purpose CPU, instead of a single purpose hash machine.

    Your question "how to recycle what will soon or later be a HUGE obsolete distributed supercomputer dedicated to SHA-256" is a tough one. Even assuming longevity of bitcoin, SHA-256 will eventually(/probably) need to be replaced at some point, which would make these rigs even more useless. I will brainstorm and sleep on this question, but currently I don't see a bright future for these particular silicon wafers.
  • PouncerPouncer Member Posts: 1
    The folks over at Nxt have figured out an innovative way to put mining rigs to good use, despite Nxt being a 100% PoS network. Miners owning SHA or Scrypt rigs can point them to this pool They will channel the hashing power to the most profitable coin to mine at any instant, and automatically convert them into NXT.
  • chris613chris613 Member Posts: 93 ✭✭
    I disagree that raiding clonecoins is an innovative or good use of hashing rigs.
  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    @Pouncer Interesting to know, but thats just a mining pool they run? They are still dependent on the existence of SHA256 PoW coins. The very model turning into Nxt is a statement against the existence of those PoW coins.

    Probably merged mining approach is possible, afaik that puts a coin in the merged chain?

    Prefer it on Ethereum, or even paid-for, Thought of this, 'Truth or Punish' aint all bad, but the bit where a large fraction has to vote to punish if you are not receiving the block in that idea is weak. Once people have the block data, the Merkle tree approach should allow any person can indicate invalid parts, and invalidate the block, take the stake.

    Really though, SHA256 for that is useless too, better go for some kind of stake..
  • HadrienHadrien MaltaMember Posts: 17 ✭✭
    @chris613‌ I slept on it as well:
    twisted thoughts:

    First, alternate use of ASIC rigs will depend on the implementation (each series are different and could permit different "alternate stuff").
    By creating smart Contracts (a bit like futures), we could organise betting on which ASIC will allow alternate use, or in other worlds, a second life cycle.
    The underlying value of those second hand ASIC would potentially increase an undetermined way.
    Skilled programmers will try to to the hack to demonstrate that one ASIC can potentially be worth more than another one.

    This creates a (speculative) market.
    Would this market grow enough, it would catch the attention of manufacturers, who logically implement ASIC with secondary use in mind.

    That's the best I could come up with for "recycling the ASIC rigs".

    ("never tell a kid who loves computer that a computer is useless ;) please")
  • chris613chris613 Member Posts: 93 ✭✭
    @Hadrien‌ betting on which ones will allow alternate use? The manufacturers already know and they just can tell us. I know too, and the answer is that unless the alternate use is ALSO finding nonces for SHA256 blockchain headers (for some other purpose, of course), the answer is no. They have no other purpose.

    Perhaps a manufacturer shipped chips with an extra feature set that was disabled and can suddenly be made available with a new rig firmware, as part of some huge publicity stunt or massive long term business gamble. I have my doubts. ;)
  • StephanTualStephanTual London, EnglandMember, Moderator Posts: 1,282 mod
    They'll make excellent paperweights.
  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    Murphies law says they're either too light or too bulky. I mean look at the usb ones dont weight nearly enough.
  • StephanTualStephanTual London, EnglandMember, Moderator Posts: 1,282 mod
    Drill a hole in that BFL Jalapeno and you got yourself one sweet ashtray :)
  • chris613chris613 Member Posts: 93 ✭✭
    They still make decent electric space heaters. Keep your feet warm under your desk.
  • zupremezupreme Chicago, ILMember Posts: 4
    The current generation of ASICs will be useless for Ethereum - at first. However there is the strong possibility that someone will write a contract or will create a "coin" on top of Etherium which leverages SHA256 and for which the originally Bitcoin-focused algorithms will eventually apply.

    Bottom line is that people (like me) who invested in mining hardware are not going to quietly let their investments turn to dust if they can avoid it.
  • HadrienHadrien MaltaMember Posts: 17 ✭✭
    @chris613‌ & @zupreme‌ :

    @vitalik‌ , ideally we have your opinion:

    So it is possible to expect something else from those rigs: "finding nonces for SHA256 blockchain headers (for some other purpose, of course)" (chris613)

    "However there is the strong possibility that someone will write a contract or will create a "coin" on top of Etherium which leverages SHA256 and for which the originally Bitcoin-focused algorithms will eventually apply." (zupreme‌)

    I am certain a smart contract can make use of SHA-256.
    Vitalik wrote that he specifically designed Ethereum with ASIC avoidance in mind (and I guess it wasn't trivial). I deduct that their is plenty of uses for the ASICs.

    one thing at a time?..

  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    @Hadrien It is essentially trivial to dodge the particularly SHA-256 ASICs.. Just dont do SHA. Lite and Doge use Scrypt. Eventually people created an ASIC for that.. So they created VertCoin, to avoid that. Not sure how effective that will be. It isnt trivial to stay ahead of potential new ASICs being designed.

    The SHA-256 ASICs are probably useless outside mining.. I mean, sure you can make your ethereum subcurrency that can be mined with them, but how do you get the subcurrency any value? If the subcurrency doesnt have value, you do not earn the energy costs to run those ASICs. Furthermore how would that value have any relation to the use of those ASICs? The community around those ASICs is essentially paying for the energy costs of running them, what would they get back for it?
  • HadrienHadrien MaltaMember Posts: 17 ✭✭
    @Jasper‌ : we are back to my first post beginning this discussion.

    Do we conclude their is no possibility to recycle the bitcoin & script ASIC rigs with Ethereum? At least for the present implementations.
    If yes that would unfortunately close the discussion. Do we agree?
  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    Yeah basically i agree.

    Only for novelty value you may be able to do general computation. Find values X so that by creating different variants of X by flipping X[i],X[j](those number bits) that H(X)[k]=f(X[i],X[j]) with f=or,and,xor,not It is novelty because it would be very inefficient, not even sure if it can be done.
  • chris613chris613 Member Posts: 93 ✭✭
    LOL @Jasper‌ - OMG man, that's one F'ing inefficient switchable logic gate you've designed. Horrific, really! In principle this is quite simple, but I'm not sure how to prove/disprove if any X exists that satisfies these functions. That would be truly neat as a college project.

    For those who don't quite understand the formal explanation - this does NOT count as a way to recycle BTC mining ASICs ;) It's basically the idea of finding a very specific input to SHA256 whose contents can be altered two bits at a time to perform basic logic operations. Going from there to "General Computation" is still a LOOOOOOONG way, and it's absurd before you even start.

  • farl4bitfarl4bit Member Posts: 2
    The miningpool at Nxt is a individual initiative and not really in line with the vision of Nxt. Proof-of-Stake is what Nxt is all about.
Sign In or Register to comment.