BitCoin inspired www alternative

[MyFirstNameIsPaul]

The BitCoin project, to my knowledge, is one of the most successful FOSS projects to date.  When I first read about it not long after its launch, I pigeonholed it into the same obscure category as Folding@home, SETI@home, etc., but BitCoin has quickly outgrown all similar projects in both popularity and computing power.  I imagine the primary driver of this success is the profit element as well as a sort of rebellious desire to buck the system.

Even on non-tech forums I see people who are basically speculators buying equipment to mine BTC.  While this may be the death knell for investors, it is surely a strong sign of the level of success of such a project.

One of the things I find unique in the BitCoin model is that producers are also consumers, so profit and production are effectively combined, but it is not a requirement that participants be producers.

I have an idea for an alternative community-based extension of the Internet that could use this profit element similarly to BitCoin.  The idea is to use some type of mesh network, such as Meraki or Roofnet, but incorporate a profit element.  Base to all communications would be a transaction of some sort of digital currency which allows the traffic to cross each node.  This way the network could propagate the same way that BitCoin has by essentially allowing non-techy types to simply purchase equipment and begin making money (as well as spending money) using the mesh.

Aside from just profit, it offers benefits such as an escape from ologopolistic ISPs; more efficient networks; financial incentive to reduce undesirable traffic; access to regions that otherwise have no access; combining of voice and data into one wireless service (where the mesh is popular enough to support the bandwidth and latency requirements); elimination of central-control and threat of ‘kill-switches’ and the like; increased difficulty in warrantless tapping; a large and profitable community that, if popular enough, could politically and even financially fight for unlicensed bandwidth.

I’m not a software engineer, so I can’t suggest the best way to create this system, however, as a marketer my instincts tell me this could be very successful.

[Stephen Gornick]

Base to all communications would be a transaction of some sort of digital currency which allows the traffic to cross each node. 

That would be something.  I wouldn't doubt there are designs out there for just that.

As far as interesting wireless projects:
  - http://fabfi.fablab.af/

[nmat]

Uhhh....  Folding@Home is processing data in the PetaFlops level... don't think Bitcoin is anywhere near that yet... good post otherwise though I guess 

Network Hashrate (Terahashs/s)
13.40

Network Hashrate (PetaFLOPs/s)
170.22

[gusti]

The BitCoin project, to my knowledge, is one of the most successful FOSS projects to date.  When I first read about it not long after its launch, I pigeonholed it into the same obscure category as Folding@home, SETI@home, etc., but BitCoin has quickly outgrown all similar projects in both popularity and computing power.

Uhhh....  Folding@Home is processing data in the PetaFlops level... don't think Bitcoin is anywhere near that yet... good post otherwise though I guess  

Folding is in the range of 10 petaflops, bitcoin is actually 170 petaflops

[bcforum]

Folding is in the range of 10 petaflops, bitcoin is actually 170 petaflops

Give that zero floating point operations are required to compute the SHA256 hashes for bitcoin, I find the number 170 petaflops slightly unrealistic.

You could perhaps make an argument for integer operations, but even that is stretching it a bit.

[gusti]

Folding is in the range of 10 petaflops, bitcoin is actually 170 petaflops

Give that zero floating point operations are required to compute the SHA256 hashes for bitcoin, I find the number 170 petaflops slightly unrealistic.

You could perhaps make an argument for integer operations, but even that is stretching it a bit.

well, I'm referencing bitcoinwatch figures, you can discuss that with Jeff Garzik
but I suspect he cannot be that wrong 

[nmat]

Give that zero floating point operations are required to compute the SHA256 hashes for bitcoin, I find the number 170 petaflops slightly unrealistic.

You could perhaps make an argument for integer operations, but even that is stretching it a bit.

True. I never really understood why did they use FLOPS to measure integer calculations...

[Xephan]

Folding is in the range of 10 petaflops, bitcoin is actually 170 petaflops

Give that zero floating point operations are required to compute the SHA256 hashes for bitcoin, I find the number 170 petaflops slightly unrealistic.

You could perhaps make an argument for integer operations, but even that is stretching it a bit.

If we just take the equipment churning out those hashes and estimate their FLOPS capability, we might get some kind of closer to reality answer.

Assuming that most of the hash rate are coming from GPU miners using the AMD GPUS either 5xxx or 6xxx series.
Based on the official SP FLOPS of these cards and what information I can dig up about hashrates at default clocks... since most miners are more keen to show off their best, it's a bit limited

The ratio of TFlops to GHash is between 6.4 to 7.5 TFLOPS per GHash/S

Going by that, 13THash would translate to about 83.2 to 97.5 TFLOPS with the real answer somewhere in middle since there's probably a mix of 5xxx and 6xxx cards, as well as some nVidia miners.

Side note, wiki says Folding is in the range of 4.1 native PFLOPS as of July 2011.

[kjj]

Folding is in the range of 10 petaflops, bitcoin is actually 170 petaflops

Give that zero floating point operations are required to compute the SHA256 hashes for bitcoin, I find the number 170 petaflops slightly unrealistic.

You could perhaps make an argument for integer operations, but even that is stretching it a bit.

If we just take the equipment churning out those hashes and estimate their FLOPS capability, we might get some kind of closer to reality answer.

Assuming that most of the hash rate are coming from GPU miners using the AMD GPUS either 5xxx or 6xxx series.
Based on the official SP FLOPS of these cards and what information I can dig up about hashrates at default clocks... since most miners are more keen to show off their best, it's a bit limited

The ratio of TFlops to GHash is between 6.4 to 7.5 TFLOPS per GHash/S

Going by that, 13THash would translate to about 83.2 to 97.5 TFLOPS with the real answer somewhere in middle since there's probably a mix of 5xxx and 6xxx cards, as well as some nVidia miners.

Side note, wiki says Folding is in the range of 4.1 native PFLOPS as of July 2011.

You missed a factor of 1000.

G=billion (10^9), T=trillion (10^12), P=quadrillion (10^15).

Assuming your ratios:
1Ghash/Sec ~= 6.4 TFLOPS  means 1 hash/Sec ~= 6,400 FLOPS.

If we switched to floating point math, we would be around 13 * 10^12 hashes/sec, which is 83 * 10^15 FLOPS, or 83 PetaFLOPS.

But we aren't doing floating work, so our FLOPS rate is still zero.

[JoelKatz]

Folding is in the range of 10 petaflops, bitcoin is actually 170 petaflops

But...  Bitcoin miners are all processing the same data, correct?t?

It would be awesome if everyone else was trying to mine bitcoins to be sent to my wallet, but alas I am the only one doing so.
All bitcoin miners are performing the same calculations, but on sufficiently different data that the results are not at all the same.

[mizerydearia]

Bitcoin for client/protocol/network
bitcoin for currency/money/value
BitCoin for erroneous confusion and inaccurate case

[Xephan]

You missed a factor of 1000.

G=billion (10^9), T=trillion (10^12), P=quadrillion (10^15).

That's what happen when I use a spreadsheet and start dividing by 1000 to make it readable and forget to label them
Thanks for catching it!

Assuming your ratios:
1Ghash/Sec ~= 6.4 TFLOPS  means 1 hash/Sec ~= 6,400 FLOPS.

If we switched to floating point math, we would be around 13 * 10^12 hashes/sec, which is 83 * 10^15 FLOPS, or 83 PetaFLOPS.

But we aren't doing floating work, so our FLOPS rate is still zero.

Well, our potential FLOPS rate would be that if for some reason the Bitcoin hash algorithm was changed to some new fangled FP based algo
So for the purpose of comparing network processing power, I think that's a fair estimate.

[MyFirstNameIsPaul]

Bitcoin for client/protocol/network
bitcoin for currency/money/value
BitCoin for erroneous confusion and inaccurate case

What do you mean?

[Xephan]

Bitcoin for client/protocol/network
bitcoin for currency/money/value
BitCoin for erroneous confusion and inaccurate case

What do you mean?

He's referring to the proper usage of the terms.

Bitcoin with the capital B refers to the client/protocol/network because it's a proper noun.
bitcoin without the capital B, unless beginning a sentence, as a common noun refers to the currency or a value denominated in bitcoin.

BitCoin is therefore inaccurate and "undefined". Although personally I'm not sure if there is a formal lexical definition for Bitcoin.