Jorge Tim贸n [ARCHIVE] on Nostr: 馃搮 Original date posted:2014-07-04 馃摑 Original message:On 7/4/14, kjj ...
馃搮 Original date posted:2014-07-04
馃摑 Original message:On 7/4/14, kjj <bitcoin-devel at jerviss.org> wrote:
> I suspect that there exist no algorithms which cannot be done better in
> an application-specific device than in a general purpose computer. And
> if there is such a thing, then it must necessarily perform best on one
> specific platform, making that platform the de facto application
> specific device.
>
> I'm not sure how one would go about proving or disproving that, but it
> seems very likely to be true.
I assumed this was obvious and self-evident for anyone who knows what
a Turing machine is, but judging from the number of smart people
wasting their time on the pursue of the "anti-ASIC" myth (also known
as pow wankery) it seems I was wrong.
Anything you can do with software you can do with hardware and
viceversa (you can even do it with ropes and fire in Minecraft!!)
Does this really need any proof?
I think it's the hard-pow cultists who have to provide a counterexample.
Very often you just need to query-replace "ASIC" with "specialized
hardware" to prove that what they're saying makes no sense.
> Keeping the algorithm simple, and ASIC-easy, has one other advantage.
> Just about anyone can sit down and design an ASIC for SHA, for example,
> leading to diversity in the marketplace. A harder algorithm can still
> be made into an ASIC (or more generally into an ASD), but will require
> more skilled designers, more expensive fabrication, etc. This actually
> concentrates the ASIC advantage into the hands of fewer people, which
> again, is contrary to the stated goals.
Yep, I think this is the strongest argument against "hard-pow".
But unfortunately you even find people that want "anti-ASIC without
being anti-GPU", which is funny because GPU just has Nvidia and AMD
and because unlike "anti-ASIC", anti-GPU is actually possible (despite
Litecoin's Scrypt having miserably failed on both accounts).
Interestingly enough, Greg Maxwell told me that the energetc advantage
of memory-hard pow ASICs is even greater than the advantage for SHA
ASICs.
Anyway, I'm working on a branch to encapsulate the proof of work that
should serve people to more easily experiment with alternate proofs on
top of bitcoind's code. I plan to use it for private chains ("proof of
signature" or "proof of script" if you prefer), and although it's not
ready yet, some people may be interested or may want to give some
feedback:
https://github.com/jtimon/bitcoin/tree/proof
I don't know if it will make it into master, but by specializing
ProofOfWork with TestnetProofOfWork we could remove
Params().AllowMinDifficultyBlocks() and its checks.
--
Jorge Tim贸n
馃摑 Original message:On 7/4/14, kjj <bitcoin-devel at jerviss.org> wrote:
> I suspect that there exist no algorithms which cannot be done better in
> an application-specific device than in a general purpose computer. And
> if there is such a thing, then it must necessarily perform best on one
> specific platform, making that platform the de facto application
> specific device.
>
> I'm not sure how one would go about proving or disproving that, but it
> seems very likely to be true.
I assumed this was obvious and self-evident for anyone who knows what
a Turing machine is, but judging from the number of smart people
wasting their time on the pursue of the "anti-ASIC" myth (also known
as pow wankery) it seems I was wrong.
Anything you can do with software you can do with hardware and
viceversa (you can even do it with ropes and fire in Minecraft!!)
Does this really need any proof?
I think it's the hard-pow cultists who have to provide a counterexample.
Very often you just need to query-replace "ASIC" with "specialized
hardware" to prove that what they're saying makes no sense.
> Keeping the algorithm simple, and ASIC-easy, has one other advantage.
> Just about anyone can sit down and design an ASIC for SHA, for example,
> leading to diversity in the marketplace. A harder algorithm can still
> be made into an ASIC (or more generally into an ASD), but will require
> more skilled designers, more expensive fabrication, etc. This actually
> concentrates the ASIC advantage into the hands of fewer people, which
> again, is contrary to the stated goals.
Yep, I think this is the strongest argument against "hard-pow".
But unfortunately you even find people that want "anti-ASIC without
being anti-GPU", which is funny because GPU just has Nvidia and AMD
and because unlike "anti-ASIC", anti-GPU is actually possible (despite
Litecoin's Scrypt having miserably failed on both accounts).
Interestingly enough, Greg Maxwell told me that the energetc advantage
of memory-hard pow ASICs is even greater than the advantage for SHA
ASICs.
Anyway, I'm working on a branch to encapsulate the proof of work that
should serve people to more easily experiment with alternate proofs on
top of bitcoind's code. I plan to use it for private chains ("proof of
signature" or "proof of script" if you prefer), and although it's not
ready yet, some people may be interested or may want to give some
feedback:
https://github.com/jtimon/bitcoin/tree/proof
I don't know if it will make it into master, but by specializing
ProofOfWork with TestnetProofOfWork we could remove
Params().AllowMinDifficultyBlocks() and its checks.
--
Jorge Tim贸n