Cost efficient mining hardware project

[cypherf0x]

FPGA clusters work, but... they're so expensive that you'll likely never recoup your investment.

The FPGA approach is the high tech way, but what if there was a way to do it simply?  The double SHA-256 is what needs to be accelerated.

There are cryptographic accelerator/security chips that will do SHA-256.  They may not be as fast as a FPGA or graphic card chip but they have a huge advantage over both in two areas

• They're cheap
• They're very low power

So what if you can only get 500KH/s out of it? It's worthless right?

Wrong, considering some of these chips cost less than a dollar and they're tiny so you can fit a whole lot of them on a board.  Yes there's IO considerations and other design requirements but those can be handled.

If you factor in the cost of power and hardware for current mining setups it's plausible to be able to create a stackable board at near the same cost point though using a fraction of the power.

And yes, this is going to be open source/open hardware.  The difficulty is going to keep going up making the power hungry GPUs less cost effective for mining.  Which means if you have the desire you can download a design and build it yourself if you so desire.

Anyone with hardware skills is welcome to join.

On fundraising, the current bitcoin stock system is crap.  On that the project will sell shares of the future project mining operations.  The shares will be PGP signed by the master project key and will later be transferable to a stock platform should a user friendly/accessible one be produced.

If you want to be part of the hardware or software team degrees are a plus but completed well documented projects are worth as much or more.

We need general support roles too.

If there is enough funding the standard development platform with be Altium as it will cover the whole process including FPGAs should one be included on board for management.  If it's on a shoestring then it'll be a hardware team vote (or whatever the majority of us already have)

There will be a lot of IC sampling and prototyping involved (Aka, the fun part) with the focus on relatively inexpensive chips already in mass production.  This is a research project meaning iterations of smaller less expensive prototypes before taking those lessons learned into making a larger board with more resource investment to loose.  I'm choosing to do it this way because of lots of experience with projects.  The big ones tend to fail, but the ones that encourage quicker iterations of working prototypes tend to survive based on the psychology of having successful attempts.

A tiger may be able to eat something big, but a colony of ants can do the same job more efficiently.

[ryepdx]

Man, I'm glad you're taking your FPGA idea in this direction. I was worried you might go rogue and screw us all over or something.

I have a bachelor's degree in computer science. I would definitely be interested in helping write software for this. Pretty much useless when it comes to hardware, though I'm interested in learning as much as I can.

[bulanula]

Man, I'm glad you're taking your FPGA idea in this direction. I was worried you might go rogue and screw us all over or something.

Really ?

Some of you really need to face the damn truth around this forum.

Bitcoin is only ( mostly ) used for people to "print" money with their computers. Some company will clearly realize this potential and join in etc. with some massive clusters and make this centralized network that has more than 50% and can be taken down by the gov ASAP ( eg this whole FPGA thing to get more than 50% of the power can even be gov. made to appear like they were not involved in taking this down eg doing it undercover LOL ) etc.

Then normal miners with GPUs only will stop using BTC as they are losing money due to electric costs and high difficulty then the whole system will collapse as this "get rich quick" scheme will not be working out for a number of reasons :

-no real value : who accepts BTC now ? NOBODY at least no major company due to it being too volatile
-people print money : always a dumb concept ; even the FED gets this wrong etc.
-even miners do not trust BTC and sell it for the cold hard $$$ as soon as they can !

The concept was good but from now on it is all going DOWNTOWN. Sad really because I got here too late to make it big from the 1-> 8 usd rate etc.

Let's face it : people who mine only mine to convert back to USD and get rich quick off BTC which will cause this to fail because nobody like Amazon offers BTC option etc.

Only option would be for everyone to get FPGAs and people to offer real services for BTC. LETS MAKE THIS WORK GUYS !

[pwnyboy]

Let's not hijack this thread with philosophical or emotional BS and keep it topical, shall we?

Count me in for hardware/software design, testing, and general input on scalability.  My background is IT/datacenter management; have already implemented designs with Microchop and Digi microcontrollers.  As an aside, you think crypto chips will yield superior results to the $16 Spartan 6's with 45k LEs?

[bulanula]

Let's not hijack this thread with philosophical or emotional BS and keep it topical, shall we?

We all know that once companies get hold of BTC they will make it go like a rotten apple. Companies do this all the time ! The end is near.

Better pull out faster than when doing it unprotected to avoid pregnancy LOL. My 5870s are on eBay already. Once they become worthless and everybody starts selling them they will become harder and harder to sell. What gamer buys old 24/7 used GPUs ? NONE.

Maybe there is future for BTC but gpu mining is all dead right now ! Don't say I did not warn you etc. !

[tachi641]

FPGA clusters work, but... they're so expensive that you'll likely never recoup your investment.

The FPGA approach is the high tech way, but what if there was a way to do it simply?  The double SHA-256 is what needs to be accelerated.

There are cryptographic accelerator/security chips that will do SHA-256.  They may not be as fast as a FPGA or graphic card chip but they have a huge advantage over both in two areas

• They're cheap
• They're very low power

So what if you can only get 500KH/s out of it? It's worthless right?

Wrong, considering some of these chips cost less than a dollar and they're tiny so you can fit a whole lot of them on a board.  Yes there's IO considerations and other design requirements but those can be handled.

If you factor in the cost of power and hardware for current mining setups it's plausible to be able to create a stackable board at near the same cost point though using a fraction of the power.

And yes, this is going to be open source/open hardware.  The difficulty is going to keep going up making the power hungry GPUs less cost effective for mining.  Which means if you have the desire you can download a design and build it yourself if you so desire.

Anyone with hardware skills is welcome to join.

On fundraising, the current bitcoin stock system is crap.  On that the project will sell shares of the future project mining operations.  The shares will be PGP signed by the master project key and will later be transferable to a stock platform should a user friendly/accessible one be produced.

If you want to be part of the hardware or software team degrees are a plus but completed well documented projects are worth as much or more.

We need general support roles too.

If there is enough funding the standard development platform with be Altium as it will cover the whole process including FPGAs should one be included on board for management.  If it's on a shoestring then it'll be a hardware team vote (or whatever the majority of us already have)

There will be a lot of IC sampling and prototyping involved (Aka, the fun part) with the focus on relatively inexpensive chips already in mass production.  This is a research project meaning iterations of smaller less expensive prototypes before taking those lessons learned into making a larger board with more resource investment to loose.  I'm choosing to do it this way because of lots of experience with projects.  The big ones tend to fail, but the ones that encourage quicker iterations of working prototypes tend to survive based on the psychology of having successful attempts.

A tiger may be able to eat something big, but a colony of ants can do the same job more efficiently.

I like this idea. I have the capital to fund the project, and this sounds fun. I'd like to talk in private if you would PM me your contact. Thanks.

[smooth]

Just thinking out loud here.  You say 500Kh/sec (maybe you made up this number?).  The low end "good" GPUs today are about 300Mh/sec.  So you would need 600 of these chips to equal one GPU.  Is that really going to be feasible in terms of h/$ and/or h/W.

[cypherf0x]

Let's not hijack this thread with philosophical or emotional BS and keep it topical, shall we?

Count me in for hardware/software design, testing, and general input on scalability.  My background is IT/datacenter management; have already implemented designs with Microchop and Digi microcontrollers.  As an aside, you think crypto chips will yield superior results to the $16 Spartan 6's with 45k LEs?

Well the idea is R&D.  There are a lot of encryption and security based chips that will do SHA-256.  It's actually prototyping boards and testing their speeds.

[bulanula]

LOLWUT they getting 200 MHs for some Spartan fpga !

[cypherf0x]

Just thinking out loud here.  You say 500Kh/sec (maybe you made up this number?).  The low end "good" GPUs today are about 300Mh/sec.  So you would need 600 of these chips to equal one GPU.  Is that really going to be feasible in terms of h/$ and/or h/W.

It's a made up number, I don't know the performance of those chips.  It may take more, but they'll still use less power.

Another alternative is to buy ATIs development hardware and make a board that uses the GPUs purely for computation.

[fabric]

Been lurking on the forums for some time now but finally registered an account so I could respond to this thread

This project sounds awesome. I'd be interested in helping out if I can. I've studied computer science and some basic comp/electrical engineering in university, work in software development (most recently in the operating system security community), and love tinkering with operating systems/overclocking/rigs.

[pwnyboy]

LOLWUT they getting 200 MHs for some Spartan fpga !

Quit trolling you fuck.  Go sell your 5870s, but do so quietly.

[bulanula]

No need for insults let's keep it clean please. No need sharing with us how you got shafted after buying your 4x5870s today etc. I knew this moment would come.

[PcChip]

I'd like to contribute somehow as well, although probably with the later stages such after the initial prototypes have been built, as I don't have as much FPGA experience as others here.

You said you might need people for "general support" roles, I'd be willing to help however I can.

I spent a few school semesters doing VHDL on a Xilinx board, so I have general working knowledge of how everything operates, I'm decent with C/C++, networking, etc - so just let me know if and how I can help out.  Maybe after you have a prototype designed and plans put out onto the internet I can download them and attempt assembling one and give you detailed feedback.

I hope this project gets going quickly and actually succeeds at making working prototypes, sounds very fun!

[chungenhung]

Let's not hijack this thread with philosophical or emotional BS and keep it topical, shall we?

We all know that once companies get hold of BTC they will make it go like a rotten apple. Companies do this all the time ! The end is near.

Better pull out faster than when doing it unprotected to avoid pregnancy LOL. My 5870s are on eBay already. Once they become worthless and everybody starts selling them they will become harder and harder to sell. What gamer buys old 24/7 used GPUs ? NONE.

Maybe there is future for BTC but gpu mining is all dead right now ! Don't say I did not warn you etc. !

Then great, please sell all your GPUs, so the difficulty can go down for the rest of us.

[chungenhung]

LOLWUT they getting 200 MHs for some Spartan fpga !

Quit trolling you fuck.  Go sell your 5870s, but do so quietly.

And, please PM me, as I will GLADLY buy the 5870s off your hands instantly.

[cypherf0x]

I'd like to contribute somehow as well, although probably with the later stages such after the initial prototypes have been built, as I don't have as much FPGA experience as others here.

You said you might need people for "general support" roles, I'd be willing to help however I can.

I spent a few school semesters doing VHDL on a Xilinx board, so I have general working knowledge of how everything operates, I'm decent with C/C++, networking, etc - so just let me know if and how I can help out.  Maybe after you have a prototype designed and plans put out onto the internet I can download them and attempt assembling one and give you detailed feedback.

I hope this project gets going quickly and actually succeeds at making working prototypes, sounds very fun!

I'm planning on trying to avoid FPGAs just because of expense though it's possible that one will end up as the control chip.

[MoonBuggy]

cypherf0x: I'm certainly interested, although not sure how much use I would be. For what it's worth, I'm currently studying for a master's in physics, focusing on computational simulation of real systems, so I've got a decent mathematics/software background, just not much on the hardware. I'm actually quite tempted to grab one of the dev boards mentioned here, not that I expect doing so to be profitable (nor entirely relevant to this project), since it seems like as good a way as any to pick up some general knowledge in an area that looks as though it could suit me well.

bulanula: Pretty much everything you say about bitcoin could equally be said of gold, surely? The fact that it can be mined and can't be traded directly for most goods and services doesn't preclude its value or usefulness.

[cypherf0x]

cypherf0x: I'm certainly interested, although not sure how much use I would be. For what it's worth, I'm currently studying for a master's in physics, focusing on computational simulation of real systems, so I've got a decent mathematics/software background, just not much on the hardware. I'm actually quite tempted to grab one of the dev boards mentioned here, not that I expect doing so to be profitable (nor entirely relevant to this project), since it seems like as good a way as any to pick up some general knowledge in an area that looks as though it could suit me well.

bulanula: Pretty much everything you say about bitcoin could equally be said of gold, surely? The fact that it can be mined and can't be traded directly for most goods and services doesn't preclude its value or usefulness.

If you want to play with FPGAs the Spartan 6 LX9 MicroBoard is a really good place to start.  You can google it, Avnet makes/sells them.  They're less than $100

[MoonBuggy]

Can't argue with 1/4 of the cost I was expecting for something to tinker with, thanks for the recommendation.