📅 Original date posted:2014-07-04
📝 Original message:Just some general comments on this topic/discussion.
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.
IO-bound is exactly the same as memory bound, for devices that have
enough memory. 20 GB is already trivial today, and you don't really get
into ask-the-wife-for-permission money until you cross 128 GB. The
exception would be if the IO was to an oracle outside of the device's
control, and artificially limited in throughput. Such a centralized
oracle would be contrary to the goals usually stated by people thinking
about anti-ASIC designs, so there isn't much point.
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.
kjj [ARCHIVE] /
npub1t9…r6cf0
2023-06-07 15:23:36
in reply to nevent1q…rn40
Author Public Key
npub1t9txg5aevth45pu9hcz075tj3ch25fydn8c38wew90qjjdjv3rqq3r6cf0Published at
2023-06-07 15:23:36Event JSON
{
"id": "c88a26f6c10e57ef5612cdafc6da1de8773a14a4d197ad1fb7493a1f6b9f4ad3",
"pubkey": "59566453b962ef5a0785be04ff51728e2eaa248d99f113bb2e2bc129364c88c0",
"created_at": 1686151416,
"kind": 1,
"tags": [
[
"e",
"a9e20ae1b6e168ee862b71e56f89b4ed7603371912b3a24e648d19074482f62c",
"",
"root"
],
[
"e",
"033dc84cb44af4ff392b7a39585a306342675c4323c2f0d1c42984e702c3fe32",
"",
"reply"
],
[
"p",
"f2c95df3766562e3b96b79a0254881c59e8639f23987846961cf55412a77f6f2"
]
],
"content": "📅 Original date posted:2014-07-04\n📝 Original message:Just some general comments on this topic/discussion.\n\nI suspect that there exist no algorithms which cannot be done better in \nan application-specific device than in a general purpose computer. And \nif there is such a thing, then it must necessarily perform best on one \nspecific platform, making that platform the de facto application \nspecific device.\n\nI'm not sure how one would go about proving or disproving that, but it \nseems very likely to be true.\n\nIO-bound is exactly the same as memory bound, for devices that have \nenough memory. 20 GB is already trivial today, and you don't really get \ninto ask-the-wife-for-permission money until you cross 128 GB. The \nexception would be if the IO was to an oracle outside of the device's \ncontrol, and artificially limited in throughput. Such a centralized \noracle would be contrary to the goals usually stated by people thinking \nabout anti-ASIC designs, so there isn't much point.\n\nKeeping the algorithm simple, and ASIC-easy, has one other advantage. \nJust about anyone can sit down and design an ASIC for SHA, for example, \nleading to diversity in the marketplace. A harder algorithm can still \nbe made into an ASIC (or more generally into an ASD), but will require \nmore skilled designers, more expensive fabrication, etc. This actually \nconcentrates the ASIC advantage into the hands of fewer people, which \nagain, is contrary to the stated goals.",
"sig": "93d69a871a4f2df11e68592ccdd2380b130fde7e366f976c3d448a1bcda5f2cfbe80a43f0c7b96d54dfb24b4f54eae9b544fc5646fe9be23204157551afcc3be"
}