Question For Bitcoin OldTimers

[ArticMine]

Re: summer 2012 - I believe Pirate's ponzi was one of the causes for both price surge and crash.
Anticipation of Dec 2012's reward halving might indeed have been another cause of the rise from 5 to 10. It also caused the hash rate to flex a bit (because of last GPU miners quitting I suppose)... but after reward halving, the doom scenario did not happen. So new January 2013 money came in and met a smaller offer than their demand.

TLDR: I don't know.

I am in the pirateat40 school of thought. During late 2011 and until August 2012, pirateat40 impacted the market by depressing the XBT/USD rate. This is typical for a massive short position. One the market realized that the pirateat40 coins were never going to hit the market because they had already being sold the stage was set for the major rally in 2013. The halving comparably had little if any impact.

[labsbitforum]

Here is my perspective which may be different than others.  A lot of the discussion in this sub-forum tends to be more "trader" dominated.  Large spectrum ranging from novice get rich quick, seasoned traders, academic traders, technical traders etc.  I'm none of those.  I understand business and markets but am not a professional trader.  I got involved for different reasons.

A. I thought it was a good idea to shake up the banking and finance industry (including irresponsible government fiscal policies).  We've watched the information age transform business, communication, information, media consumption, etc.  Banking has been highly controversial over the last 2000 years.  Here are a couple favorite perspectives from leaders who championed the best of American ideals.

“The Government should create, issue, and circulate all the currency and credits needed to satisfy the spending power of the Government and the buying power of consumers. By the adoption of these principles, the taxpayers will be saved immense sums of interest. Money will cease to be master and become the servant of humanity.”
- Abraham Lincoln

"The real truth of the matter is, as you and I know, that a financial element in the large centers has owned the government of the U.S. since the days of Andrew Jackson."
- U.S. President Franklin D. Roosevelt in a letter written Nov. 21, 1933 to Colonel E. Mandell House

B.  I have always been a computer enthusiast.  I've built more than 100 computers and it was enjoyable to push another frontier with CPU and GPU mining.

My perspective on the halving.  It has a large impact on mining.  It will obsolete a ton of gear in an instant.  We are already headed to greater and greater centralization of mining and this will add to it.  Centralization was not at all the intent of bitcoin and if not addressed this momentum will reform around a 2.0 digital currency that is decentralized and probably has other core innovations.

[notme]

My perspective on the halving.  It has a large impact on mining.  It will obsolete a ton of gear in an instant.  We are already headed to greater and greater centralization of mining and this will add to it.  Centralization was not at all the intent of bitcoin and if not addressed this momentum will reform around a 2.0 digital currency that is decentralized and probably has other core innovations.

I disagree about centralization being inevitable in Bitcoin mining, but let's assume it is true.  How would you change mining to make it more resistant to centralization?

[labsbitforum]

My perspective on the halving.  It has a large impact on mining.  It will obsolete a ton of gear in an instant.  We are already headed to greater and greater centralization of mining and this will add to it.  Centralization was not at all the intent of bitcoin and if not addressed this momentum will reform around a 2.0 digital currency that is decentralized and probably has other core innovations.

I disagree about centralization being inevitable in Bitcoin mining, but let's assume it is true.  How would you change mining to make it more resistant to centralization?

Ideally the "peers" should be CPU based.  That enables broader participation and access.  The trend to more and more specialized equipment GPUs --> FPGA --> ASICs leads directly to the network being more commercialized.  Like any industry, the companies working in this space will compete and this typically evolves to a monopoly or oligopoly.

The title of Satoshi's PDF is "Bitcoin: A Peer-to-Peer Electronic Cash System".

We are quickly moving to the point where mining is controlled by a few big players.  Increasingly you see the ASIC server farms being collocated in data center type facilities in locations with very inexpensive power.  

That's not peer-to-peer

[notme]

My perspective on the halving.  It has a large impact on mining.  It will obsolete a ton of gear in an instant.  We are already headed to greater and greater centralization of mining and this will add to it.  Centralization was not at all the intent of bitcoin and if not addressed this momentum will reform around a 2.0 digital currency that is decentralized and probably has other core innovations.

I disagree about centralization being inevitable in Bitcoin mining, but let's assume it is true.  How would you change mining to make it more resistant to centralization?

Ideally the "peers" should be CPU based.  That enables broader participation and access.  The trend to more and more specialized equipment GPUs --> FPGA --> ASICs leads directly to the network being more commercialized.  Like any industry, the companies working in this space will compete and this typically evolves to a monopoly or oligopoly.

The title of Satoshi's PDF is "Bitcoin: A Peer-to-Peer Electronic Cash System".

We are quickly moving to the point where mining is controlled by a few big players.  Increasingly you see the ASIC server farms being collocated in data center type facilities in locations with very inexpensive power.  

That's not peer-to-peer

It is impossible to make something that a general purpose CPU can do better than a custom chip, so try again.  Well, impossible is a strong word, but I'm going to need a plausible explanation of how you are going to achieve such a feat.

[gog1]

halving will drive prices up!!!

[labsbitforum]

It is impossible to make something that a general purpose CPU can do better than a custom chip, so try again.

A highly customized chip can do 1 thing better.  If you adapt what it needs to do beyond its original design it becomes worthless.

[notme]

It is impossible to make something that a general purpose CPU can do better than a custom chip, so try again.

A highly customized chip can do 1 thing better.  If you adapt what it needs to do beyond its original design it becomes worthless.

This is true.  So how does this fact help you design a coin that resists custom chips unless you change the algorithm all the time with no preplanning of what is coming next?  That may be a possibility, but the decision as to what to use next would need to take place publicly so insiders can't take advantage and build custom chips.  There are also common operations in nearly all hashing algorithms, so a custom chip that could handle multiple algorithms could probably still outdo a CPU or GPU.

[Syke]

Ideally the "peers" should be CPU based.  That enables broader participation and access.  The trend to more and more specialized equipment GPUs --> FPGA --> ASICs leads directly to the network being more commercialized.

Hogwash. It doesn't matter what hardware is used for mining. If mining is profitable, then big farms will be built. That's how people think. If you can make $1 per day mining on a desktop CPU, then a farm will be built to mine $1,000,000 a day on a huge farm of desktop CPUs. We saw it in the early days of mining when botnets were redirected to mining purposes.

[notme]

Ideally the "peers" should be CPU based.  That enables broader participation and access.  The trend to more and more specialized equipment GPUs --> FPGA --> ASICs leads directly to the network being more commercialized.

Hogwash. It doesn't matter what hardware is used for mining. If mining is profitable, then big farms will be built. That's how people think. If you can make $1 per day mining on a desktop CPU, then a farm will be built to mine $1,000,000 a day on a huge farm of desktop CPUs. We saw it in the early days of mining when botnets were redirected to mining purposes.

Also this.  So again, I ask, how are you going to stop centralization?  The important thing is not that we keep the big players out, it's that anybody who want to can be a player and that even if you aren't a player, you can verify things aren't wonky with just a CPU.

[labsbitforum]

Ideally the "peers" should be CPU based.  That enables broader participation and access.  The trend to more and more specialized equipment GPUs --> FPGA --> ASICs leads directly to the network being more commercialized.

Hogwash. It doesn't matter what hardware is used for mining. If mining is profitable, then big farms will be built. That's how people think. If you can make $1 per day mining on a desktop CPU, then a farm will be built to mine $1,000,000 a day on a huge farm of desktop CPUs. We saw it in the early days of mining when botnets were redirected to mining purposes.

Also this.  So again, I ask, how are you going to stop centralization?  The important thing is not that we keep the big players out, it's that anybody who want to can be a player and that even if you aren't a player, you can verify things aren't wonky with just a CPU.

True, I also recall admins at schools and businesses getting in trouble for CPU mining on their machines without permission as well.  I dont claim to have a technical solution on exactly how to make it more "peer-to-peer".  I'm sure it could be done.  I think a lot of people active in theses forums have a lot invested in bitcoin and thus tend to be quickly dismissive of anything negative.  I really like bitcoin.  If I didn't I wouldn't be here.  I'm bringing fourth my perspective to generate constructive dialog on an aspect of bitcoin that I believe is stuck on a path that will cause more centralization.

It seems pretty obvious to me that mining is becoming more centralized.  Give me some reasons you think its not and wont naturally evolve to that?

[notme]

Ideally the "peers" should be CPU based.  That enables broader participation and access.  The trend to more and more specialized equipment GPUs --> FPGA --> ASICs leads directly to the network being more commercialized.

Hogwash. It doesn't matter what hardware is used for mining. If mining is profitable, then big farms will be built. That's how people think. If you can make $1 per day mining on a desktop CPU, then a farm will be built to mine $1,000,000 a day on a huge farm of desktop CPUs. We saw it in the early days of mining when botnets were redirected to mining purposes.

Also this.  So again, I ask, how are you going to stop centralization?  The important thing is not that we keep the big players out, it's that anybody who want to can be a player and that even if you aren't a player, you can verify things aren't wonky with just a CPU.

True, I also recall admins at schools and businesses getting in trouble for CPU mining on their machines without permission as well.  I dont claim to have a technical solution on exactly how to make it more "peer-to-peer".  I'm sure it could be done.  I think a lot of people active in theses forums have a lot invested in bitcoin and thus tend to be quickly dismissive of anything negative.  I really like bitcoin.  If I didn't I wouldn't be here.  I'm bringing fourth my perspective to generate constructive dialog on an aspect of bitcoin that I believe is stuck on a path that will cause more centralization.

It seems pretty obvious to me that mining is becoming more centralized.  Give me some reasons you think its not and wont naturally evolve to that?

I'm not dismissive because I'm heavily invested... my holdings are tiny.  I'm dismissive because I've been involved on a very technical level since 2010 and I have training and work experience building computer hardware.

Mining has a tendency to centralization today because mining hardware is not commodity hardware.  It is produced by very few people, and those close to the manufacturers are the ones who have the large operations.  Once the hardware manufacture catches up with the top of the line manufacturing processes used by CPU and GPU manufactures, there will no longer super-moore's law gains in efficiency and the margins will be squeezed out of mining hardware production.  At this point, there will be nobody with an advantage and mining will once again decentralize.  It is just a speed bump, not a "new normal".

[labsbitforum]

I'm not dismissive because I'm heavily invested... my holdings are tiny.  I'm dismissive because I've been involved on a very technical level since 2010 and I have training and work experience building computer hardware.

Mining has a tendency to centralization today because mining hardware is not commodity hardware.  It is produced by very few people, and those close to the manufacturers are the ones who have the large operations.  Once the hardware manufacture catches up with the top of the line manufacturing processes used by CPU and GPU manufactures, there will no longer super-moore's law gains in efficiency and the margins will be squeezed out of mining hardware production.  At this point, there will be nobody with an advantage and mining will once again decentralize.  It is just a speed bump, not a "new normal".

Thanks for sharing your perspective.  I like a lot of it and I hope you are right.  But, even if you are right that the hardware becomes commoditized and readily available there are still economies of scale that will make it more profitable to be bigger and drive those with lots of capital to make bigger operations in areas with very low power costs.  The effect of that would be driving out participation by anyone not living in areas with artificially low power costs.  Wouldn't it?

[notme]

I'm not dismissive because I'm heavily invested... my holdings are tiny.  I'm dismissive because I've been involved on a very technical level since 2010 and I have training and work experience building computer hardware.

Mining has a tendency to centralization today because mining hardware is not commodity hardware.  It is produced by very few people, and those close to the manufacturers are the ones who have the large operations.  Once the hardware manufacture catches up with the top of the line manufacturing processes used by CPU and GPU manufactures, there will no longer super-moore's law gains in efficiency and the margins will be squeezed out of mining hardware production.  At this point, there will be nobody with an advantage and mining will once again decentralize.  It is just a speed bump, not a "new normal".

Thanks for sharing your perspective.  I like a lot of it and I hope you are right.  But, even if you are right that the hardware becomes commoditized and readily available there are still economies of scale that will make it more profitable to be bigger and drive those with lots of capital to make bigger operations in areas with very low power costs.  The effect of that would be driving out participation by anyone not living in areas with artificially low power costs.  Wouldn't it?

Yes, mining will tend to cluster where power is cheapest, but other factors are in play.  For example, in the apartment I am currently renting, I have electric heat.  Mining hardware is just as efficient as an electric heater, so for the 6 months a year I need extra heat, running mining hardware is essentially free.  Other than power costs, and perhaps saving a small amount per unit on shipping 100 vs 1 unit there isn't much economy of scale involved in mining.

[Scottsdale]

Those days where awesome. I was playing around when price was 13$, and watching the price increase 8-25% everyday for about 4 months was the funnest thing I could imagine. It's like, well I made another 3 grand today, oh another 4 grand, okay im rich