5 ms·

>So yeah, there's a huge amount of R&D and competition to get the best stuff out there. And Bitmain dominates. You're right that Bitmain dominates ASIC Bitcoin

by tmh88j 9y ago

>So yeah, there's a huge amount of R&D and competition to get the best stuff out there. And Bitmain dominates. You're right that Bitmain dominates ASIC Bitcoin miners, but why? Surely there are more than enough capable engineers to design ASIC miners in the US and some VC's willing to fund them, or is there something I'm missing? Manufacturing costs would be high in the US, but it's not like foreign manufacturing is an unknown concept to US based companies.

wmf 9y ago
There have been a ton of mining ASIC companies but all but two of them went out of business. For some reason VCs have mostly refused to touch mining so the ASIC vendors had to use crowdfunding which went horribly wrong.
RhodesianHunter 9y ago
Just a hunch: No-one in the industry with the means to do anything about it thinks the trend is sustainable enough to invest in a 6mo-1yr setup followed by another year before profit?
- tmh88j 9y ago
  Peter Thiel is throwing his weight into Bitcoin. Surely there are more people out there like him, albeit less vocal about BTC/cryptocurrency.
  - wmf 9y ago
    It sounds like he and many others are investing in BTC itself, not the Bitcoin ecosystem. If you can get a 10x return per year just by hodling, why bother to develop anything? Even selling shovels isn't that profitable. When people say a (hyper)deflationary currency discourages investment, this is what it looks like.
    - tmh88j 9y ago
      >It sounds like he and many others are investing in BTC itself, not the Bitcoin ecosystem. If you can get a 10x return per year just by hodling, why bother to develop anything? I completely agree as of 2018, but Bitcoin has been around for almost 10 years now and has seen "real" returns for the past 4 or 5 years. I suppose my question should be restated as "Did Bitmain take a gamble and it paid off?" I still find it strange that not a single prominent US based BTC miner manufacturer has hit the scene.
      - wmf 9y ago
        Butterfly Labs, HashFast, and CoinTerra were all US-based. Then you had KnCMiner in Sweden, Spondoolies in Israel, etc. I would say that Bitmain was a huge gamble in the early days since the ASICs depreciate very fast and thus any delay in developing them can be fatal. But after selling eight generations of hardware at very high margins they can probably coast now.
        tmh88j 9y ago
        >Butterfly Labs, HashFast, and CoinTerra were all US-based. Then you had KnCMiner in Sweden, Spondoolies in Israel, etc. None of them produced anything. CoinTerra was the only that didn't close as a result of charges related to fraudulent activity. Not to be a jerk, but that's not what I was asking. You provided supporting evidence to my claim. >I would say that Bitmain was a huge gamble in the early days since the ASICs depreciate very fast and thus any delay in developing them can be fatal. But after selling eight generations of hardware at very high margins they can probably coast now. That's more in line with what I asked. I'm not in disagreement, but rather looking for any research into why the mining market currently stands this way.
  - prostoalex 9y ago
    Risk tolerance. With net worth of $2.6 billion Peter Thiel could lose $26 million and make it back with a 1% fluctuation of the market. Other people less so.
    - tmh88j 9y ago
      I responded to the question what was asked. >Just a hunch: No-one in the industry with the means to do anything about it thinks the trend is sustainable enough to invest in a 6mo-1yr setup followed by another year before profit?
  - soundwave106 9y ago
    Aren't ASICs extremely expensive to develp? The one StackExchange article I found on ASIC pricing (https://electronics.stackexchange.com/questions/7042/how-much-does-it-cost-to-have-a-custom-asic-made https://electronics.stackexchange.com/questions/7042/how-muc...) seems to confirm that. My impression is there's a significant amount in the investing class who see Bitcoin as a hyped bubble. Sure, there's Peter Thiels, but there's also people like Warren Buffett who won't touch it with a ten foot pole at the moment. (https://www.theguardian.com/technology/2018/jan/10/bitcoin-and-cryptocurrencies-will-come-to-a-bad-end-says-warren-buffett https://www.theguardian.com/technology/2018/jan/10/bitcoin-a...). You don't need ASICs per se for general blockchain technology... the main advantage of ASICs are speed of mining, something that probably won't be needed as much once the general mania pops one way or another. So an ASIC mining chip seems to me fairly vulnerable. For instance, if you just spent a million making a lot of ASIC mining chips, news like China possibly starting to crack down on mining (https://www.ft.com/content/adfe7858-f4f9-11e7-88f7-5465a6ce1a00 https://www.ft.com/content/adfe7858-f4f9-11e7-88f7-5465a6ce1...) would probably not make you happy.
solotronics 9y ago
I think it has to do with what is essentially money laundering, they buy miners with their Chinese money to mine coins to move out of China or hide.
prostoalex 9y ago
Volatility of returns and guaranteed increased complexity? I think 21.co got AH financing on the idea of building a "miner for the masses" https://www.coindesk.com/21-intel-bitcoin-mining-strategy/ https://www.coindesk.com/21-intel-bitcoin-mining-strategy/ but they pivoted since then. They could probably provide a more informed comment.
icelancer 9y ago
The startup time is significant and VC money doesn't go into BTC mining. Furthermore, the engineers almost certainly have to live in China and have engineering / fabrication experience there. And you can virtually guarantee your first product will be terrible and obsoleted by Bitmain nearly immediately. Manufacturing costs in the US wouldn't just be high, they would be impossible to actually get done. MOQ on something like this in American factories wouldn't be in the 100-500 range, probably not even 1,000. American factories are extremely non-agile, non-risk taking, and generally do not serve startups very well. BTC being over $10k is a very new thing, too. The insane mining profits at a 1MB blocksize + $10k+ price may not last and is a new signal. It takes a lot of time, effort, luck, and perseverance - and Jihan is a formidable enemy. Sometimes it's best to just avoid a high-risk area where many, many others failed.

by tmh88j 9y ago

wmf 9y ago

There have been a ton of mining ASIC companies but all but two of them went out of business. For some reason VCs have mostly refused to touch mining so the ASIC vendors had to use crowdfunding which went horribly wrong.

RhodesianHunter 9y ago

Just a hunch: No-one in the industry with the means to do anything about it thinks the trend is sustainable enough to invest in a 6mo-1yr setup followed by another year before profit?

tmh88j 9y ago
Peter Thiel is throwing his weight into Bitcoin. Surely there are more people out there like him, albeit less vocal about BTC/cryptocurrency.
- wmf 9y ago
  It sounds like he and many others are investing in BTC itself, not the Bitcoin ecosystem. If you can get a 10x return per year just by hodling, why bother to develop anything? Even selling shovels isn't that profitable. When people say a (hyper)deflationary currency discourages investment, this is what it looks like.
  - tmh88j 9y ago
    >It sounds like he and many others are investing in BTC itself, not the Bitcoin ecosystem. If you can get a 10x return per year just by hodling, why bother to develop anything? I completely agree as of 2018, but Bitcoin has been around for almost 10 years now and has seen "real" returns for the past 4 or 5 years. I suppose my question should be restated as "Did Bitmain take a gamble and it paid off?" I still find it strange that not a single prominent US based BTC miner manufacturer has hit the scene.
    - wmf 9y ago
      Butterfly Labs, HashFast, and CoinTerra were all US-based. Then you had KnCMiner in Sweden, Spondoolies in Israel, etc. I would say that Bitmain was a huge gamble in the early days since the ASICs depreciate very fast and thus any delay in developing them can be fatal. But after selling eight generations of hardware at very high margins they can probably coast now.
      - tmh88j 9y ago
        >Butterfly Labs, HashFast, and CoinTerra were all US-based. Then you had KnCMiner in Sweden, Spondoolies in Israel, etc. None of them produced anything. CoinTerra was the only that didn't close as a result of charges related to fraudulent activity. Not to be a jerk, but that's not what I was asking. You provided supporting evidence to my claim. >I would say that Bitmain was a huge gamble in the early days since the ASICs depreciate very fast and thus any delay in developing them can be fatal. But after selling eight generations of hardware at very high margins they can probably coast now. That's more in line with what I asked. I'm not in disagreement, but rather looking for any research into why the mining market currently stands this way.
- prostoalex 9y ago
  Risk tolerance. With net worth of $2.6 billion Peter Thiel could lose $26 million and make it back with a 1% fluctuation of the market. Other people less so.
  - tmh88j 9y ago
    I responded to the question what was asked. >Just a hunch: No-one in the industry with the means to do anything about it thinks the trend is sustainable enough to invest in a 6mo-1yr setup followed by another year before profit?
- soundwave106 9y ago
  Aren't ASICs extremely expensive to develp? The one StackExchange article I found on ASIC pricing (https://electronics.stackexchange.com/questions/7042/how-much-does-it-cost-to-have-a-custom-asic-made https://electronics.stackexchange.com/questions/7042/how-muc...) seems to confirm that. My impression is there's a significant amount in the investing class who see Bitcoin as a hyped bubble. Sure, there's Peter Thiels, but there's also people like Warren Buffett who won't touch it with a ten foot pole at the moment. (https://www.theguardian.com/technology/2018/jan/10/bitcoin-and-cryptocurrencies-will-come-to-a-bad-end-says-warren-buffett https://www.theguardian.com/technology/2018/jan/10/bitcoin-a...). You don't need ASICs per se for general blockchain technology... the main advantage of ASICs are speed of mining, something that probably won't be needed as much once the general mania pops one way or another. So an ASIC mining chip seems to me fairly vulnerable. For instance, if you just spent a million making a lot of ASIC mining chips, news like China possibly starting to crack down on mining (https://www.ft.com/content/adfe7858-f4f9-11e7-88f7-5465a6ce1a00 https://www.ft.com/content/adfe7858-f4f9-11e7-88f7-5465a6ce1...) would probably not make you happy.

solotronics 9y ago

I think it has to do with what is essentially money laundering, they buy miners with their Chinese money to mine coins to move out of China or hide.

prostoalex 9y ago

Volatility of returns and guaranteed increased complexity? I think 21.co got AH financing on the idea of building a "miner for the masses" https://www.coindesk.com/21-intel-bitcoin-mining-strategy/ https://www.coindesk.com/21-intel-bitcoin-mining-strategy/ but they pivoted since then. They could probably provide a more informed comment.

icelancer 9y ago

The startup time is significant and VC money doesn't go into BTC mining. Furthermore, the engineers almost certainly have to live in China and have engineering / fabrication experience there. And you can virtually guarantee your first product will be terrible and obsoleted by Bitmain nearly immediately. Manufacturing costs in the US wouldn't just be high, they would be impossible to actually get done. MOQ on something like this in American factories wouldn't be in the 100-500 range, probably not even 1,000. American factories are extremely non-agile, non-risk taking, and generally do not serve startups very well. BTC being over $10k is a very new thing, too. The insane mining profits at a 1MB blocksize + $10k+ price may not last and is a new signal. It takes a lot of time, effort, luck, and perseverance - and Jihan is a formidable enemy. Sometimes it's best to just avoid a high-risk area where many, many others failed.