False negatives are a problem, but as the percentage of them is less than the gain in total amount of hashes, it's still a net win to do approximations.
Read the paper. The savings come from smaller/faster hardware due to approximate carry propagation in the hashing cores. It has nothing to do with submitting false positives. False positives as easily weeded out by a regular CPU in less time than it takes to prepare a block and send it over the network.
The real cost to the provider is to have the bandwidth operational for transferring 100MB. Whether you do that in 3 days or 30 is not really relevant to the discussion.
You make it sound like it was obvious, but SHA-256 hashing requires 64 rounds of a hash function, and bitcoin mining requires a double SHA-256 hash. Even a small error in the hash function is likely to propagate quickly to the point where practically all of the results are bad. So, in this case the approximation must be nearly perfect to be worthwhile.
The waste is composed of a mix of radioactive materials with widely varying half lives. Some half lives, such as Strontium-94 are only 75 seconds, which means that it'll be completely gone in a few months, and radiation/heat produced by the waste will drop. Another common waste product Sr-90 has a half life of 30 years, but there are also waste products with half lives of millions of years. Most of the danger comes from the parts with intermediate half lives, as they produce lots of radiation, and will do so for years.
As long as they keep it close, because the stuff that we call "spent fuel" still has 99% of the original energy locked inside. At some point, we'll want to dig it back up and actually use it.
Industry use of gold is very small, and does not explain its value. And the monetary value is a big reason for people to buy gold jewelry (otherwise they'd all get gold plated stuff).
No, the false positives are easy to detect locally. A regular CPU in your PC could detect a false positive in much less than a microsecond, and the number of false positives would be small to start with.
No, that part of the paper is bullshit. Once you think you've found a hash, it's easy to verify with a regular CPU before you'd claim to have found a block.
I agree it won't hurt for people to spend some time brainstorming about this in their spare time. But before we actually set up a budget to look into this on a more structured scale, it would probably be better to use the same budget to improve road safety, for example, or encourage people to stop smoking.
Slashdot Mirror
If you want to say it's shockingly bad, first establish what a proper percentage should be
The proper percentage should be at least 99%, but I'll settle for 99% among members of the government, and 90% for the public.
What's stopping you from just minding your own business
Self interest, of course. But if you're such a proponent of minding your own business, why are you arguing here ?
Or perhaps, two different surveys were held, each asking only one of the two questions.
Other people's choices are not your choices.
Other people's choices affect my life and health, so I prefer that these are well informed.
You can have knowledge of something you don't believe.
It matters for making better informed choices about antibiotics, vaccinations, stem cell research, and various other topics.
politically-correct "belief" in evolution
There's nothing political about it. Just cold hard science. And the same applies to global warming.
The 75% number about elephants is still shockingly bad.
False negatives are a problem, but as the percentage of them is less than the gain in total amount of hashes, it's still a net win to do approximations.
Read the paper. The savings come from smaller/faster hardware due to approximate carry propagation in the hashing cores. It has nothing to do with submitting false positives. False positives as easily weeded out by a regular CPU in less time than it takes to prepare a block and send it over the network.
The real cost to the provider is to have the bandwidth operational for transferring 100MB. Whether you do that in 3 days or 30 is not really relevant to the discussion.
Current above-ground gold reserves are big enough for more than 650 years of industrial consumption. The scarcity is purely artificial.
You make it sound like it was obvious, but SHA-256 hashing requires 64 rounds of a hash function, and bitcoin mining requires a double SHA-256 hash. Even a small error in the hash function is likely to propagate quickly to the point where practically all of the results are bad. So, in this case the approximation must be nearly perfect to be worthwhile.
The waste is composed of a mix of radioactive materials with widely varying half lives. Some half lives, such as Strontium-94 are only 75 seconds, which means that it'll be completely gone in a few months, and radiation/heat produced by the waste will drop. Another common waste product Sr-90 has a half life of 30 years, but there are also waste products with half lives of millions of years. Most of the danger comes from the parts with intermediate half lives, as they produce lots of radiation, and will do so for years.
Where I live, there's a food home delivery service that accepts bitcoin as payment.
As long as they keep it close, because the stuff that we call "spent fuel" still has 99% of the original energy locked inside. At some point, we'll want to dig it back up and actually use it.
All you are showing is that bitcoin, money, cars, and houses have value.
You misunderstood. There are plenty of uses, but the combined total of industrial use is only 10% of the new gold mined every year.
Not true. Right now, finding a block requires doing an average 7E20 hashes before you find a good one. Verifying only takes one hash.
Industry use of gold is very small, and does not explain its value. And the monetary value is a big reason for people to buy gold jewelry (otherwise they'd all get gold plated stuff).
No, the false positives are easy to detect locally. A regular CPU in your PC could detect a false positive in much less than a microsecond, and the number of false positives would be small to start with.
No, that part of the paper is bullshit. Once you think you've found a hash, it's easy to verify with a regular CPU before you'd claim to have found a block.
It is true that mining is only worthwhile if you have the cheapest equipment and lowest electricity rates. The same is true for many other endeavours.
Increasing the albedo makes the photons bounce back, which requires a bigger change in momentum than just stopping them.
I agree it won't hurt for people to spend some time brainstorming about this in their spare time. But before we actually set up a budget to look into this on a more structured scale, it would probably be better to use the same budget to improve road safety, for example, or encourage people to stop smoking.