Bitcoin Forum

So I discovered Bitcoin today and soon after that dove into my usual obsessive research mode about mining.  The difficulty of generating bitcoins is going to keep going up and the amount of power the current mining rigs I've seen is going to go up.  I'm a computer engineer by trade so I'm kind of a bastard of electrical engineering and computer science.  I do know FPGAs though.  I've got a PICO EX-300 board in my workstation with 16 Xilinx Spartan FPGAs on it.  I still have a lot of work to do with optimization of the chips but I've got a single core running at about 117MH/s with a fairly quick and crude design.

I've been using the board to crack WPA 2 at some pretty amazing speeds.  If I can put enough time in it I should have a core design that could be loaded on USB/PCIE FPGA boards that'll use a lot less power and have some serious hashing power.

Looking forward to seeing the final results.

Give us the averaged Mhash/power consumption stats for all 16 Xilinx, including the price for one of them, then we can talk.

The performance/watt isn't an issue, the cost however will.

The challenge will be to find an inexpensive FPGA board with the right performance/cost.

I can design a mining board with FPGAs and that's what I may end up doing but that's not trivial and would take some time.

You did all this in a single day?

I have a compiler that'll convert C code to VHDL that's why I said it was crude but it works as a proof of concept.  I've just got to go through the code and write properly now, which may take a bit depending on my workload.

Highly interested in this. How much would you say the FPGA's for usb/pci-e cost?

http://www.picocomputing.com/E101/ (http://www.picocomputing.com/E101/)

They're about $400

It could be a lot cheaper.  If I get the VHDL optimized I can get it turned into an ASIC and put a number of those on a PCIe express card/other format card.

Development costs/time vs job will slow me down but I'm not going to be a tool and post my bitcoin address or a donate button like a beggar.  I could also get distracted by some other shiny project that has a greater return for my efforts.  That or it works so well I just keep it for myself and profit.

I may play with the idea of a standalone miner that you just put the configuration on a micro-sd card add power and Ethernet and get mining.

I doubt that this could bit a single 5770 card, but i am looking forward for your results :)

I know it can.  I've used FPGAs for brute forcing crypto before and it was a lot more involved than hashing.  Graphics cards aren't made for hashing. They're good at it because of the way their processing is optimized (ATIs over Nvidia), now with an FPGA you can get a processor that does everything in hardware, no software or code to run to perform the algorithms.  That's why you see very high performance out of FPGAs since they've essentially become a processor that does that one task purely in hardware.

I've prototyped a dedicated AES cypto chip for an in-line encryption device that was running at 10Gb/s with an FPGA.  It'll work, but I don't have a fat government contract funding development.  I'm shooting for the most cost effective solution in power an cost as possible.

That's the challenge, and I like challenges.

if the algorithm can be implemented in a hardware, the power consumption will be decreased a lot, will this cause the BTC price down?

The BTC price is not tied to power prices (currently it costs ~ 1USD in electricity to mine 1 BTC!).

Should there be an easily available solution for a relatively cheap ASIC box with a few GHashes/s and low power consumption, the difficulty would probably skyrocket and the electricity cost on GPUs would go so high that it becomes as ineffective as CPU mining today.

I guess it'll still take a while (as ASICs also don't materialize out of thin air!) but I personally expect this to happen this year still, should bitcoin continue to be so popular.

Already a 400USD FPGA with (if you optimize the code maybe) ~150 MH/s that has a laughable power consumption might be a danger for GPU mining in the medium term (as they would be very easy to cool)

So looking over the FPGA I realized how much of the chip was unused.  Why is this important?

FPGAs are reconfigurable, don't think of them as a single processor.  I did some coding and added another hashing pipeline to the chip and smoothed out some bumps.  A single pipeline is now doing about 133MH/s with the chip around 210MH/s total (a bit less for the crude pipeline I cut/pasted/mauled).  This is significant because as long as I have enough enough logic gates available I can add more pipelines.  How many more?  Quite a few  ;D

If I put this on a Virtex-7 which is designed for 400Gb/s operations and took advantage of it's large number of gates then it would be a hashing monster,

Don't ruin bitcoin-.-

It wouldn't be too nice to have spent 1000$ when a 400$ chip will do it 40 times better-.-

The concept appears viable but I fear might be cost prohibitive
Gpu's can be used for some other purpose once mining becomes in efficient
What's a person to do with the left over $1000 worth of fpga equipment in a yr or 2

Continue to mine with extremely electricity efficient hardware.

Apologies but no more development information will be posted.  I've been offered a 25% share from someone that owns 2 FPGA clusters.  If you haven't seen that type of hardware before think a 156 FPGAs per machine.

* sigh *

well, i guess we know where mining is headed...

I had a feeling this is what would happen.

Can you at least give us the rough specs of the clusters, so I can ballpark about how many Gh/sec you'll end up pulling?  If they're machines like this (http://www.dinigroup.com/new/DN2076k10.html), then you're probably going to be pulling anywhere from 500Gh/sec to 5.5 Th/sec per machine.  If it's a small cluster of perhaps 24 machines used for cryptographic attacks and research, then that one cluster will probably end up with more compute power than the entire rest of the network. 

Man, now I want to dig out my old Spartan 3 fpga board and see if I can get a working hash engine out of it just for giggles.



For people who don't know a ton about FPGA stuff, you can run a hash engine natively in hardware using a hew hundred gates at most for MD5, and run one hash per clock cycle, with a clock speed of 5-550 Mhz.  Most of these chips have anywhere from 2000 gates for the barest of bare bones $50 DIY kits to about 9 million on the latest Virtex 7. 

Assuming the hash engine took 350 gates (2 orders of magnitude reliable), the chip runs at 550 MHz, and has 9 million gates, and was just 20% efficient, that is 80% of the time the chip is waiting for data or or otherwise not hashing, you could see about 2.8 trillion hashes per second.  TRILLION HASHES PER SECOND.  Even using worst case figures, a 5 million gate device at 250 Mhz, 5000 gates per hash engine, and a 5% utilization factor, you still get 50 Gh/sec, over 200 times the hash calculating power of a Ati Radeon 5850.

Think more along the lines of this http://www.dinigroup.com/new/DNBFC_S12_12_Cluster.html (http://www.dinigroup.com/new/DNBFC_S12_12_Cluster.html)

I was afraid this would happen.  I don't think it's good for this stuff to be kept secret.  To that end:

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

I will offer a 10 BTC bounty to the first person who either:

1) Makes publicly available the source code and complete setup instructions for an FPGA-based Bitcoin mining device that can be created with off-the-shelf hardware.

or

2) Offers an ASIC-based Bitcoin mining device for sale to the general Bitcoin community, and makes publicly available all schematics, source code, and other relevant information used to develop said device.

Either of the above devices must cost no more than $0.60 USD per megahash per second that they provide, and must consume no more than 0.3 watts per megahash.

Designs must not be encumbered by any sort of "intellectual property" restrictions, with the exception of GPL-style copyleft licenses.

- -Chris Acheson, 5/17/11
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
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=onT1
-----END PGP SIGNATURE-----

It's not a whole lot, but hopefully others will join me.

Your figures are wildly off.  A ununrolled miner takes about 90K LU, and a internally pipelined one that can reach high clockrates is probably more like 180K LU.  Millions of gates.   This pushes you into the most expensive FPGAs currently available just to get good performance...

FPGAs are great for H/j  but not H/$ at the moment. The upcoming generation of FPGAs may improve this somewhat.  The number's you're describing aren't really realistic except via fully custom asic with NREs in the million dollar range.

can you opensource the code?

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


I will gladly match Chris Acheson's bounty of 10BTC on his terms.

A basic FPGA miner isn't a lot of work and it would be a fun project
for someone who hasn't done this kind of work before.  A _fast_
fpga miner which will achieve competitive performance would be a decent
accomplishment.

I think it's important to the health, security, and public confidence
in bit coin that a few large private parties do not retain a substantial
long term advantage in their ability to control the hashchain.

Making sure that the public has the lowest cost access to the mining
state of the art should be helpful for this purpose.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)

iEYEARECAAYFAk3SruYACgkQrIWTYrBBO/p58wCfYeGrfT9ptb/bOapN0zJ0Dt9J
XFoAoMuISTUBUGqeOqJc1TBesyG6e0Fh
=WAwq
-----END PGP SIGNATURE-----

You probably have to make that bounty about 1000 times higher to make it remotely interesting to those who have the know how and money to make this stuff in the first place.

The high end Spartan-6 has ~150K gates.

Is this type of thing cost effective for miners to buy just for mining? Nope you'd likely never pay off cost of the cluster from your mining.
Is it cost effective for someone who already owns units used for other work? Very, considering each chip only pulls ~5 watts and they're sitting idle.


The bounty is laughable... A person keeping the code to themselves could profit a lot more than that and keep the competitive advantage.  You may not like it but it's capitalism and just a good business practice to keep that advantage.

Ok, buisiness... I have agreements not to release my current design which involves FPGAs

142MeH5qrYmHBEGkzVYGMKaLc66KVHZe4q

If I get enough to that address to fund getting an ASIC produced, I will along with an easy to use board.

Well I stand corrected, and was hilariously off base on my values.  This is still a pretty awesome business plan to run on idle FPGA clusters that would otherwise sit there doing nothing.  The marginal cost is pretty much zero since the hardware was purchased for something else and the depreciation expense is borne by someone else.

Does anyone else have this kind of knowledge? I'd definitely be interested in buying some sort of PCIe card with onboard mining-optimized ASICs if it means better Mh per slot.

Yes. 150K LU  is not enough to fit an unrolled engine with internally pipelined adders. It's enough to fit _one_ unrolled engine, which you'd probably be lucky to get running at 100MHz (=100MH/s).     Maybe you could do some awesome stunts, depending on the platform and somehow get two in, though I don't see it.

If it's otherwise idle capacity, then fine— it would be profitable. But you're not talking about a huge competitive advantage for anyone yet, certainly not a huge short term competitive advantage.


Much more and it simply becomes easier to write it myself.  It's quite simple to write a SHA2-256 engine in verilog, though harder to get it going fast.

To someone who knows the tools and has the development kit handy, it's probably two days of work to get something basic going, though the skys is the limit on optimizations.  I'd offer the use of hardware, but the largest programmable FPGA I own is only 27K LU, which is too small to be interesting for this.

The reality is that someone will eventually do it for love or money and reality trumps capitalism.

Moreover, shaking confidence in the security of bitcoin by securing a large private advantage wouldn't be economically sensible for anyone developing this stuff in any case.

I would put much more hope in the upcoming 22nm southern islands GPUs. Even if there are no architecture changes at all, theoretically performance per watt ratios could reach up to 4x the current 69xx ones. We might reach 10000 SPs in a dual GPU.

Custom ASICs are retarded expensive.  I know the NSA/Some universities got them for DES/Triple DES encryption key breaking, but they cost a few million to develop, and were never in large enough demand to make the non recoverable engineering costs low enough on a per chip basis.

It's a bit easier now, though still not very cheap.  You can hand the team FPGA files and they'll turn it into an ASIC.  The only way to make it cost effective per chip though is to get a run of 10k+ or so or tooling costs will kill you.

Yeah, if 5 years from now you had several companies doing the block generation as a business model where they collect the tips, then you would probably see a big rackmount box packed to the brim with ASIC logic in order to drive the per transaction cost down low enough to still make a profit.

I just realised that Bitcoins future depends on using an algorithm that is not possible to put in hardware like this. If it is, there will probably only be one mining company left after a while because of the economy of scale.

That's the problem with cryptographic money systems.  If you can do it in software you can do it in hardware faster.

What performance coule one expect from these devices:

http://cgi.ebay.com/Altera-Cyclone-EP1C6-NIOSII-EP1C6T144C8-FPGA-Dev-board-/120724210798

http://cgi.ebay.com/Cyclone-III-FPGA-EP3C25-Board-USB-Blaster-12864LCD-/260735361173

http://cgi.ebay.com/Altera-Cyclone-NIOS-II-FPGA-Development-Board-EP2C8Q208-/250809160224

http://cgi.ebay.com/Cyclone-II-FPGA-EP2C8-board-/280447719711

http://cgi.ebay.com/Altera-Cyclone-NIOS-II-FPGA-Board-ByteBlaster-2C35-VGA-/160487569726

I'm not looking to finance the entire project myself.  If I were, I'd just hire an engineer and not bother talking about it here.  I'm trying to get the ball rolling and prompt others who are also concerned about Bitcoin remaining a distributed currency to chip in.


Really?  We should give you money so that maybe something will come of it?

Never used it but it has enough gates that it should be able to run a few pipelines

I would like to buy a FPGA and program it with LabVIEW. Is this a good idea?

So someone should release the code and maybe get a bounty?  You can play with maybe all day.  In the end I already have a working prototype and now someone else with more FPGA experience than myself to polish the code.  He develops chips for a living, I develop hardware boards and embedded software so that seems like a pretty reasonable combo for getting something done.

You can, but LabVIEW + LabVIEW FPGA module + supported FPGA dev board is about $5700

Hello cypherf0x! Good to see a fellow FPGA developer on the forums  ;D

I'm a bit curious about your numbers. Working with an Altera Cyclone3 and Cyclone4, which are the brothers of the Xilinx Spartan series, I've only ever seen 1MHash/s per 1K LUTs. The chips you describe have the equivalent of 75K LUTs (Xilinx has a funky design), which means no more than 75MHash/s. If you are indeed getting >200MHash/s out of a single Spartan 75K, that would be quite wonderful!

Anyway, thank you for posting your findings and, even if you don't post anything more about your research, it's always good to have a new, knowledgeable member on the forums.

The chips on the boards have about 100k LUTs  23k slices with 4 LUTs/slice
Try adding parallel pipelines

If you're serious about this, you should arrange to have the contributions put in escrow until you actually release something.  Just putting up a black-hole donation address means that no one knows how much has been contributed, and if the total only gets halfway there the whole thing's a loss.

Anyway, the bounty isn't aimed at you specifically, so I'm going to split it off into its own thread.

For which platform? The PICO EX-300 platform? You're probably talking about a platform with Spartan-6's on it, because those do indeed carry four 4-LUTs per slice, with the LX150 totaling ~150k 4-LUTs.


I develop on a C120, and I have different designs, some of which are indeed pipelined. The pipelined designs get 1Mash/s per 1K LUTs.

Perhaps the Xilinx devices can pack far more bang-per-LUT than Altera's for SHA-256 designs? I shall certainly investigate. Again, thank you for sharing your numbers.

I got only one question. Where the heck can you even buy these things like "PICO EX-300?" All I find are specs and specs. I guess you need to study computer engineering for 10 years just to see one of those?

For anyone skeptical about FPGA abilities the video below is for MD5 hashing but it's the same principle.

http://www.youtube.com/watch?v=zEwWvVP_RU0

They're expensive enough you have to call the sales office to order them.

In your haste to create faster miners, be careful that you don't destroy that which you seek.


My understanding is that there is a bigger problem, in that if one person/organisation controls over 50% of the Bitcoin network, then it is effectively compromised and bitcoins will no longer be a viable e-currency.

See this thread: http://forum.bitcoin.org/index.php?topic=8653.0 (http://forum.bitcoin.org/index.php?topic=8653.0)

https://en.bitcoin.it/wiki/Weaknesses#Attacker_has_a_lot_of_computing_power (https://en.bitcoin.it/wiki/Weaknesses#Attacker_has_a_lot_of_computing_power)

Attacker has a lot of computing power
An attacker that controls more than 50% of the network's computing power can, for the time that he is in control, exclude and modify the ordering of transactions. This allows him to:
Reverse transactions that he sends while he's in control
Prevent some or all transactions from gaining any confirmations
Prevent some or all other generators from getting any generations
The attacker can't:
Reverse other people's transactions
Prevent transactions from being sent at all (they'll show as 0/unconfirmed)
Change the number of coins generated per block
Create coins out of thin air
Send coins that never belonged to him
It's much more difficult to change historical blocks, and it becomes exponentially more difficult the further back you go. As above, changing historical blocks only allows you to exclude and change the ordering of transactions. It's impossible to change blocks created before the last checkpoint.
Since this attack doesn't permit all that much power over the network, it is expected that no one will attempt it. A profit-seeking person will always gain more by just following the rules, and even someone trying to destroy the system will probably find other attacks more attractive. However, if this attack is successfully executed, it will be difficult or impossible to "untangle" the mess created -- any changes the attacker makes might become permanent.

If anyone is looking for an inexpensive FPGA to experiment with try the SPARTAN-6 LX9 MICROBOARD.  I've gotten a lot of messages asking about it and these boards are USB and cost less than $100

Without having an actual Spartan-6 LX150 board on hand, I ran my design through ISE quickly. This showed that the LUT consumption is indeed similar to Altera's, so there does not appear to be any area improvements by using a Xilinx device over Altera.

What I do not know, however, is how fast Spartan-6 LUTs operate compared to Altera's, for apples-to-apples speed grades. If they run faster, it would indeed be possible to get more bang for your LUT. I get 80MHz in my design, resulting in 80MHash/s burning 80K LUTs. The Cyclone4-150 or Spartan6 LX150 may fit two full hashing pipelines (128 SHA-256 rounds per full hashing pipeline). This would double their performance. The Cyclone4-150 achieving 160MHash/s. If the Spartan6 is faster, it could possibly achieve >200MHash/s as you've reported.

You could get faster speed grades, but those are typically a bit more expensive. I haven't calculated whether a fast speed grade would balance out the cost for its improved hashing speeds.

Trying to make any sense of this.
a) You have a 120+ stage unrolled pipelined engine at 133MHz. You fit 1.58 of em? what the hell is 0.58 of a engine?
b) You have a single registered round running at 133MHz. one bitcoinhash = double-sha256 takes 128 or so clocks. you fit 200 of those - ~ 208Mh/s.
let's assume B
you need to store at least a..h and W 0..15, that's 24*32 = 768 FFs per engine.
times 200 engines. thats 153600 FFs
a S3-5000 has 66560 FFs... nope
a S6 LX100 has 126576 FFs... still nope
a S6 LX150 has 184304 FFs... 83% utilization just for the storage FFs. far edge of plausible

For adder utilization it gets hilarious, you need at least 8 32-bit adders per round.
Times 200 single-round engines thats 1600 32bit adders...
half of a S6s slices have carry logic, each of those can do 4 bits of a adder, that's a max of 988 32 bit adders on a S6 LX100, 1439 on a LX150... we need 1600... ?!?

I have the sneaking suspicion someone didn't realize one bitcoinhash = 2 sha256 blocks...

It's actually about 90MH/s over time per pipeline but the speed average jumps around a bit at first but settles over a longer run.

Okay, so now you're fitting 2 pipelined engines on a LX150.
need 120 rounds, thanks to cheating with W updates etc you can get it down to ~6 32 bit adders per round avg, times 120 ... 720 or so 32 bit adders per engine, 1440 adders.
So *only* a bit over 100% slice utilization of a LX150, just for the adders. Yeah, sure.

I don't know where you came up with 133MHz out of MH/s.  There is the 'about' and 'around' meaning values are not absolute.  The speed average was a bit high initially.  You're also making design assumptions.  There are highly optimized commercial hashing cores available for FPGAs too.

Okay, so you fit "around" 1.5 engines on a chip. is it me or doesn't that make any sense at all?
Edit:
Yes, I make assumptions about sha256. it's sha256. the round function including W update needs at least 8 32 bit adders. no amount of "optimizing" changes that.
And those "highly optimized" commercial cores? barely 120MHz on a S6, 65+ clocks/block, and you can maybe fit 70 on a LX150. 65Mh/s wooo...

I never said I fit 1.5 engines on a chip.  I apologies if some of the numbers implied that since they were ballpark estimations based on short runs.

You're free to doubt, it's your time spent.

You can actually fit 1.5 engines on a chip, assuming an engine is a full 128 rounds of SHA-256 (that's 128 because you need to do it twice to get the final hash that Bitcoin expects). One full engine, at 128 rounds, and a second half-engine, at 64 rounds, with a mux in front to switch between processing new data and finishing old data.

I've considered doing that for my C120, which would fit one full engine in 80K LEs, and the half-engine in 40K (if I'm lucky). Or just get my hands on a C150 and try desperately to cram two full engines on it :P

Hey cypherf0x,

I just got into bitcoin and ran across your post here. I have my own startup and we run a small design/assembly service as part of our business. We have the capability to assemble FPBGA/LGA parts on PCBs and I would be really interested in working with you on a low cost Spartan-6 FPGA board. I know I would be willing to put ups some of my own cash to fund some initial board revisions, and with a little help from the community we might be able to produce a batch of these at a really compelling price.

Interested?

Aaron

Yeah, send me a PM with your email.

How much does a Cyclone4-150 board cost?

Also do you get a discount of you buy the Spartan boards in bulk?

Kinda shame you can't get more hashes out of the Spartan boards cause compared to say a cheap 5850 it gets killed. If these boards cost like $50 or something that may be worth it for 80 Mhash/s I'd definitely buy 4 to begin with - maybe more.

I think you'll get best results if you can share the workload between the GPU and the FPGA. I don't have the technical know how to get this going but it might be possible to do it with something like the ODG1 in your mining rigs PCI slot.

http://www.linuxfund.org/projects/ogd1/

I might be interested in helping out with this project with either time or money.  I'm a system software engineer with some experience writing VHDL with a EE degree.

I did my own back-of-the-envelope calculation similar to others on this forum.  I figured you could probably fit a single unrolled pipeline in a LX75 Spartan 6.  I agree with cypherf0x that it's not completely clear whether it's possible or not.  For the sake of argument, let's say it is.  Then if you could run at 100 MHz then you could get 100 MH/s.

The problem is that it doesn't seem much cheaper than GPU computing.  The Spartan 6 LX75s are about $100 each at Avnet.com (one of Xilinx's distributors).  By the time you figured in the fabrication and assembly costs you might break even in terms of "hashes per capital expense"

However, it would certainly be more power-efficient... and a lot more dense.  If you were clever you could put arrays of them on DIMM-like modules like this design: http://www.sciengines.com/copacobana/ .  You could potentially fit 100 or more FPGAs in the case that a single desktop computer fits in.

100 FPGAs @ 100Mhz = 10Gh/s in one computer case? I would gladly pay for that. My biggest constraint right now is physical space, followed distantly by the amount of electricity my circuit breaker can route my way.

Do want. O_O

Given that these numbers are right:

Do you think it would be worth it? If you're willing to do this, then probably a few others are too. If you buy a rig like this for US$50000 (dunno the price, just assuming something), you'd need to make ~7142 bitcoins at a value of US$7 each before it's worth it, or 143 blocks. Remember that the network only hands out 210000 chunks until 2013, so difficulty will increase as you are doing this. The question is - will it pay off? Will you actually be able to use this rig to get those 143 blocks?

Maybe smaller investments is the way to go for bitcoins (unless you already have some fancy hardware). Buy something, generate a couple of blocks on it, then buy some new hardware that is better, and generate a few blocks on that, etc... Heavy investments will probably be overtaken by bitcoin limitations and technology before they've paid off. If so, what would be the right level of investments? If someone works on it, I guess it should be possible to find an optimal price level vs. block performance ratio that could be a buyers guide to get into bitcoins.

Unless you want to gamle that bitcoin prices will increase a lot, of course. :) You're already gambling that they'll stay at current levels.

I think if you want to do this make it a public mining rig. e.g. people can buy shares in it to allocate the capital for the purchase of the mining and FPGA hardware then people will get a return on these shares. The alternative is is a few greedy guys getting all the bitcoins for themselves and then everyone else  loses interest in them and moves onto the next craze. I would also like to suggest that people can be paid in shares rather than bitcoins directly for their help, elbow grease, attention and expertise they put towards helping build the company. Please also make everything such as the business plan and technical documentation and diagrams open source.

+1 to everything you just said. I'm not an engineer, but I can supply some capital and some programming help if some enterprising and learned person wants to point me in the right direction.

ryepdx: thank you. I don't have much of anything to contribute myself except a few wild ideas but I may buy some shares if I like the proposal. It could happen on Nafario's http://glbse.com/

It could be something that allows people to diversify their investments and actually trade the shares which may be rising even at times when bitcoins are flat or in a correction. Then they can hold, use or sell their dividends i.e. the bitcoins.

GLBSE is a good idea, I’d probably buy shares.

$50k is WAY too much for a 10Mh/s rig, IMO. Just right for a 30Mh/s to 50+Mh/s rig, though.


Not if one gets in early enough. If an investment is to pay off, it doesn't matter if you've invested $2k or $20k. The number of days it'll take to break even stays the same. It's just a matter of how much risk you are willing to take on.


That would be interesting...

I like the idea of this.  :)
Now what can we do to make it happen ?

Well I would suggest we make a list of the steps that companies normally go through for an IPO and take out all the boring and expensive regulation compliance. There would need to be a sort of prospectus and a business plan and cash flow forecasting. There would need to be break even analysis and profit and loss statements. These could be done in wiki form there would need to be other software as well possibly google docs like spreadsheets or modified versions of SMB Ledger.

Despite all the hardwork it would be a very exciting thing to be involved in.

You would deploy the miners in multiple locations around the globe so there would be reduced risk of natural disaster bringing down the company and you would make a special mining pool for these miners and backup servers in different locations for the mining pool server as well. You could even make it more p2p in nature so that any miner could be upgraded to become a pool server if need be.

You should also plan ahead so that more shares could be sold to raise more capital for further expansion and upgrades for example things even better than FPGAs http://www.technologyreview.com/computing/37406/?p1=A1&a=f. The idea being that the open and public FPGA miner grows bigger and spreads far more tentacles than any secret and private FPGA miner.

To stop the company from getting over a 50% share everything is open source so that others could make a similar company.

Well, I have to point out that the $100 cost I pointed out is just for the FPGAs themselves.

You've got to have someone draw up board schematics, lay out a board, fabricate it, possibly spin it if there are hardware design bugs, assemble the prototypes, debug it.  You still need some kind of host system for it.

By the time you do all that you're looking at probably $20000 instead of $10000 for a 10 Gigahash system.  Some of it could be made up in volume by building 10 or more such systems... but some of it can't.  Assembly/fabrication costs are always going to be large for small quantities.  So now you're looking at a $/hash that's very close to PCs with ATI cards.

You'll also need time to get this running.  If you started today and did it right, you could be running in a couple of months... at which point the bitcoin difficulty will be quite a bit higher.

If T hash rigs start to appear it will send GPU mining the way of CPU mining and essentially destroy a little of the spirit of bitcoin, at the moment its open to all and the reward for participating is actually a very good incentive to participate even if you have just 1 card.

 Large FPGA clusters basically as stated skew the shift of power so much that people will not participate and the reward goes to the few as opposed to the many,

Also in my opinion this idea of a uber mining company is exactly what bitcoin wanted to avoid that not one organization controls the network or the flow of money, you become shareholders essentially in a bank and your proposal is a world bitcoin bank!

PS where do i signup .... here are my 30 silver bitcoins.

That's still within the range of what I'd be willing to pay for a 10Gh/s machine, especially if the machine had the same footprint as a single tower. I would prefer this over 10Gh/s with GPU machines due to the space savings.


How is this different from when GPUs sent CPUs the way of the dinosaur? If FPGA rigs remain modular, such that one could buy a 1Gh/s pluggable card for around $1,500 to $2,000, there really wouldn't be any difference from the situation we have now. Except this time the money would remain in the Bitcoin community rather than going to ATI or NVidia.

At any rate, I imagine the rise of mining corporations, or at best cooperatives, is inevitable. The lone hunter gives way to society. At best we can prepare for when it happens.

Nice counter argument, i think the major difference is that CPUS/GPUS are common place, the FPGA is a bit specialised and intially it will be limited to those with the technical knowledge and the capital to go down that road.  Where as most can stretch to purchasing a gpu with a fair mhash rate.

I hear what your saying, we all knew FPGAs are out there, i guess i just feel the the short golden age of community powered bitcoin will drift away and it will be a battle of the corporations opposed to a battle between individuals.

Not a lot of "corporations" will be interested to invest in a market that small, volatile and unpredictable as bitcoin...

Most likely FPGA software will be released, a few people jump on it, drive difficulty to heights that are not fine for GPU miners anymore and the whole system collapses. As soon as some try to get back on board, someone just flips the switch on an FPGA cluster and it's over again.

People will lose interest in mining and - to be honest - it seems like the single most attractive thing about Bitcoin IS mining, not having any other benefit of a cryptocurrency. Even miners here seem rather to cash out at any(!) rate than holding onto their earnings for some time.

I don't think so. It's not too different from the GPU stuff I guess. The same fears. As long are enough people doing FPGA stuff there won't be a crash or something. I also think sooner or later someone will publish an open source version. Of course there will be people doing it from early on, but in order to not have the prices go down and to prevent people from losing interest one simply needs competition.

yeah but everyone had a gpu to start with  ;) and it means corporations like the one proposed in this thread or co-ops perhaps for a lack of a better word, would control the bitcoin network.

Well i guess we see what happens.

There may be some options. I am just getting ideas here but there is a company that offers low cost FPGA's: http://www.xess.com/

It appears that the older X1800 ATI cards used an FPGA to enable crossfire. You may be able to hack together a configuration where some of the workload from your late model card is offloaded to FPGA running a custom ROM on the X1800 ATI card. These cards seem to be listed for about $30 on ebay. No idea if this will work or even be worthwhile but it might be a step in the right direction.

http://techreport.com/articles.x/9138/2

There are more recent cards with like the PICO EX-160  that could also be used and potentially in combination with a GPU:

http://www.siliconwolves.net/frames/projects.html


There is also an intel CPU with integrated FPGA the Atom E600C series:
http://www.thinq.co.uk/2010/11/22/intel-launches-fpga-equipped-atom/

http://www.mardhikaputra.com/intel-introduces-new-processors-that-can-atom-e600c-configurable.html

And then there's this:

http://www.xtremedata.com/products/accelerators/pci-express/xdpcie3000

They also have some in socket (co-processor) accelerators but they are for older sockets.

This idea may have been discussed elsewhere, but what if someone produced a purely combinational logic circuit for calculating SHA-256 hashes. Such a circuit would be relatively small (low power consumption, easy to keep cool, easy to develop) and ridiculously fast, since you would get outputs as fast as you give it inputs. However, you'd need to develop one circuit for hashing the hash of the block header plus the nonce and another for hashing that hash (different input lengths). This isn't an issue if both circuits are combined, though it does double the initial computations needed for designing the circuit. Am I making some unreasonable assumptions here or would it be possible to achieve a multi-GH/s device?

based on the huge difficulty jump today.... I think someone has succeeded in fpga/asic bitcoin flooding :(

You still need clocking.  See here. (http://en.wikipedia.org/wiki/Propagation_delay)

The idea of an ASIC (or FPGA, which is like an ASIC without the AS part) is to implement the logic such that all processing for each round is done in parallel, thus one round per clock.  You can then add pipelining to get an effective throughput of multiple rounds per clock.

That seems to be a common misconception present in this thread.

FPGAs aren't powerful. They're actually quite inefficient and slow  :P Read ArtForz's posts for a good breakdown and debunking.

Someone with free access to a huge stack of FPGAs isn't any better off than someone who has free access to a huge array of GPUs, as far as hashing power is concerned.

As for an ASIC, I highly doubt anyone is foolish enough to drop >$1,000,000 of their own cash to have developed one. Also, it takes a long time from start to finish. Bitcoins weren't all that popular a year ago. They still aren't (outside of our little bubble).

Suppose someone was foolish enough to do so. Well, sucks for them. Good luck getting $1,000,000 out of the Bitcoin market right now. The only feasible profit would be from selling them and spreading the risk to the customers.

I post occasional updates on my progress with FPGA mining in this thread. (http://forum.bitcoin.org/index.php?topic=2362.0) If I succeed in making FPGA mining feasible, I fully intend to release a product.

I also still have on my TODO list the release of an FPGA Mining Reference Design, open-source.

I hope this helps clear out the mis-information and fear. As always, I am happy to answer questions about my project, FPGA mining, and FPGAs in general. It's an exciting topic for me and I hope my knowledge can be of some use to the community.

FPGAMiner, what else would explain the difficulty jump?  I'm thinking I ordered my 5850's at exactly the wrong time.

The goal is to have each block come out in 10 minutes, on average.

The difficulty target is increased every 2016 blocks.  So, if hashing power is growing, the blocks just before a retarget will be coming in much too quickly.  If the retarget shoots for a 10 minute average based on the pool at the time, before long we'll be spitting them out in seconds.

So, you calculate the rate of growth and extrapolate it out a week.  The first 1008 blocks take too long, and the second 1008 blocks are too quick.  Hopefully, if you guessed right about the pool growth, you get it more or less right overall.

get me right, I know nothing of fpga's as they apply to mining but I can imagine a few implementations and also the potential for optimization . . .

http://upload.wikimedia.org/wikipedia/commons/0/07/Xilinx_S6-SP601_board.jpg

More people rushing in.

I'm sure a lot of people ordered hardware back when the exchange rate first jumped. Over the past week they've probably been receiving their hardware, configuring machines, and bringing their mining online. Meanwhile, popularity continues to increase.

We've all known that difficulty has been climbing at a rapid rate for quite some time now. These jumps shouldn't come as a surprise to anyone.

As for FPGA mining, it's been debunked many times already. FPGAs are good at what they do, but they're not magical.

GPUs, on the other hand, are really, really, really good at crunching lots of numbers in parallel. And they're relatively cheap.

yeah well we just get the fpga to manage the arrays

Neat idea! An FPGA driven board, with PCI-e controllers and an ARM core in the FPGA. You could then run ribbon cable to quite a number of video cards, depending on how many PCI-e cores fit into the FPGA. Luckily we only need PCI-e 1x, which seems do-able with an FPGA.

My back of the napkins calculations couldn't justify the price, though. Perhaps for large mining farms. The majority of GPU mining cost comes from the cards and PSUs. Such a solution would only cut out the motherboard, CPU, and RAM, which really aren't that much. Not to mention how difficult it would be to get fglrx drivers running on such a stripped down system.

What we really need to do is buy the GPU processors in bulk directly from AMD and build massive GPU array boards  :P

EDIT:
Personally, I don't find the difficulty jump surprising. But to answer your question, it was the recent spikes in the market price, compounded by the naturally growing number of miners. Newcomers see the market skyrocketing like that and they all start firing up the machines they have lying around. We've gotten lots of new traffic over the past week or two from Slashdot, Reddit, and I think an article or two.

I know I just brought on an extra 3GH/s online this past week or so. Also take a look at how many posts we have about people trying to get their new rigs up and running, or looking to buy one.

To be honest, the difficulty increase isn't that bad. It's quite appropriate for the market price, which jumped from ~$3-$4 all the way up to $10 not too long ago. Regardless of the difficulty increase I'm going to break-even on my hardware in a timely fashion just as I had hoped. Probably before schedule!

EDIT2:
Almost forgot! The phatk kernel was released shortly after the difficulty jump before last (jump to 157K). This increased everyone's performance by 10MH/s or so. That was about a 3% increase for most people, which most certainly contributed to this larger difficulty increase.

Those of you voicing concerns about the possibility of Bitcoin being taken over by mining corporations with FPGA/ASIC clusters should check out the bounty thread (http://forum.bitcoin.org/index.php?topic=8713.0) for an open-source FPGA mining implementation.  Encouraging people to develop this stuff out in the open will help to prevent any one party from gaining too much of an advantage.

You've pretty much answered your own question. There are probably a thousand other aspiring miners in the exact same position. It happens after every large increase in price, and combined with a Slashdot front page it's a disaster for difficulty.

So my 1,000 dollar 1Ghs system isn't a total waste.  Fwwweeewww.

*bell ring*
We have a winner!

Two FPGA clusters went live.

I can't really describe the feeling of learning that the two initial units were for prototyping and there were RACKS of them to be used when idle.

Very true, but what if there was already a source of something equivalent to high speed SHA-256 ASIC. ;)

Does anyone here think the original creator(s) of BitCoin ever forsaw this happening?
If not, how do you think they feel about it?

There are SHA-256 ASICs around, but they are basically I/O limited. They are fine for getting a hash of some given string, but in a Bitcoin miner you are incrementing a part of the input in a specific way. This can be done inside the chip, so you don't need a particularly high I/O with the outside world.

I'm curious about that our 'mystery miner' bribes cypherf0x for not posting developement details, even if the main reason stopping the public to start FPGA mining seems to be the high price of the FPGAs (and a low hash rate compared to the high price). Even availability of some source code wouldn't make a difference at all because most people won't spend a bazillion dollars on FPGA clusters. Obviously mystery miner doesn't want us to know something. Maybe there's a quite affordable way to do fpga mining.
Does someone know the price to buy/rent copacobana?  ;D

I, for one, welcome our new miner overlords.

This is the beginning of the end for bitcoins. Those who control the mining control the coins. If you're a big enough fan you'll probably be able to convince yourself that it's still better than normal fiat money, but the average person won't.

It's not like FPGAs are some secret inaccessible technology. They are just expensive. Know what else is expensive? Four 6990s.

Why?  Just because mining won't be as effective for the average person doesn't mean that it cannot still be used as a currency.

I think it's a reference to the >50% computational power attacks on the network

Sure, as long as you accept any demands they set to have your transfers be included in the official block chain. Blocks that are found by other miners can just be ignored, because they will be voted out eventually.

Ah, gotcha.  Yes, that would be a problem.

Yes the problem is not the lower profits of the average miners, it's the high centralication of hashing power. And if some universities or whoevers fpga cluster can put this much hashing power to the network, some 3 letter agencies can put bitcoin down with the snap of a finger (ok, not REALLY putting it down but performing the >50% hashing power attack, thats not much better).
What bitcoin needs to get strong enough is that the average miner is using the same maximum effective technology which any entity could bring up. Therefore miners must switch to fpga/asic mining or the algorithm needs to be extended to be no more highly parallel computable (but I don't know if that's even possible as every software basically is faster if implemented in hardware).

Didn't we already determine that fpga's aren't any faster than GPU's?

I don't know, I think we had different opinions about his.

They don't control mine.

They control the transactions that you might want to make with your bitcoins.

Now for a picture to help you feel better:
http://ryepdx.com/dune-cat.jpg

Seriously though, we need to get this tech into the hands of the rest of the community ASAP.

Did you really intend to hotlink a hotlinking notice?

And no miners do not control my transactions, because all that they can do is refuse to process them, and there are thousands of others who will not.

Wow, I'm sorry for getting off-topic, but completely disagree. If these people even exist, the technology exists, and if the technology exists, the idea exists, and if there's an incentive to implement this idea, then there will be competition among all.

Do you think people are just going to sit idly by with this newfound information, and all the people "in-the-know" are going to huddle together and laugh maliciously at taking over bitcoin? Hell no, there going to start competing amongst themselves! And let them! Let bitcoin go huge, let people put big money into it! Why? Because they'll be the ones with the biggest interest to advance it, to market it, and to incorporate it into a reasonable currency.

No, people always go a little crazy when power appears to be consolidating, but don't underestimate the free market my friend. We're all competitive, and I can only hope that someday, engineers are being hired to develop for bitcoin, research bitcoin mining, and market for bitcoin.

EDIT: I love what I see here. Lots of smart people huddling together attempting to open-source the new idea. I'm continuing on my GPU mining route (as a programmer), but I may become obsolete by the time I finish! That's technology for you!
EDIT2: Ok, some of you now have me concerned. :tinhat: I still don't expect 50%+ ownership by any single entity however

if that kind of mining power gets utilized it will completely destroy the BTC

edit...

sorry the terahash / sec that was quoted earlier...

It would be trivial for them to do as well.  Once you get >%50 of the computational power you can basically counterfeit transactions and as long as > 50% of the network validates it, it's valid to the protocol.

I still find absolutely no evidence of a massive hashing machine, whether powered by unicorn dust or by FPGAs.

One person has been making very sketchy claims, which of course he can't provide any evidence for because the Men In Black offered him 25% to keep quiet.  That is 25% of a venture that is highly unlikely to be profitable even if they didn't have to pay him hush money, by the way.

If you think that the recent jump in difficulty was caused by a mythical super-cluster coming online in the last day or two, you really need to go back and read up on the retargeting process.

No, no, no.  This is completely 100% WRONG.  You don't seem to understand how any of this works at all.

With more than 50% of the hashing power, you can hide a double spend for a few minutes, and only from funds that you have legitimately.  You can not counterfeit anything.

It would basically show that you had several confirmations before the network finally rejected it, right?

Only if you could get several blocks ahead of the rest of the network. If someone waits for a couple confirmations you have to have 3 blocks in reserve with the double spend in them to beat it.

Actually it was 25% and an NDA on the implementation.

If I did a screencast of the system then someone could say it was faked with a script too.  In the end there are always people who don't believe despite whatever you show them.

I stand corrected then, I haven't really put time into researching attack vectors yet.

So you did bring a cluster online?  What sort of GH/s is this cluster putting out?

Are you sure they are on?  I can't see a noticeable bump in total network power.  The increase is steady within normal variance.  Perhaps a lot of GPU miners shut down their systems at the exact same time?  Sure you didn't mix hash/s and Mhash/s?

Wrong. If only one person or entity utilizes that kind of power, that could destroy the bitcoin. What we need is a public service to lower the barrier to entry for people. If FPGA miners are as cost-effective as the Th/s figure quoted earlier, then it would probably be quite profitable for a person to invest in an FPGA mining business. In the long run, it's really just the cost of keeping a cluster running that matters most, assuming you can make back your initial investment.


I second the question.

[edit]


True. But could you humor us anyway? Your claims would have a bit more veracity to them if you provided evidence. As they say on teh interwebz: pics, or it didn't happen.

Okay, so the most likely existing FPGA user with this kind of hardware is either a crypto set-up or banking/financial house that used it for algo trading.

So I'm gonna have a guess that it was an existing financial entity of some sort ... which means they will have an incentive to set-up a decent bitcoin exchange with easy interface with existing currency markets to go along with their mining operation. Basically a 'rogue' from the current oligopolies is breaking ranks and going to "have a go". If this is the case, it is great news.

Edit: oy yeah, if there is one finance house out there ready to mobilise their existing FPGAs to spin a bitcoin profit there will be dozens ... a lot of these guys are "on the bones of their arse broke" after the financial collapse of fiat money, just hanging out for every QE injection from the bernank.

Okay, yes. But it's very, very hard good news to handle for those of us who just dropped a couple grand on mining equipment only a few days ago. I want to at least make back my investment before FPGA mining makes it all worthless!  :'(

I warned earlier ... can't find the post, about getting too heavily into GPUs because of threat they would go the way of CPU when FPGA came on-line ... scrapped my GPU expansion plans out of sheer nerves last week ...

EDIT: here it is
http://forum.bitcoin.org/index.php?topic=6903.msg101253#msg101253

I second that. But AFAIK this will not be massive anytime soon, so maybe we still have time to recover our investement :)

I actually view it as a GOOD thing if an FPGA cluster was what the big hashing increase was due to.  Hopefully, that means that the difficulty rate increase will slow down, unless/until more FPGA clusters make their way online.

If it's just due to a bunch of us buying more GPU's, then there's no reason that the difficulty increases won't continue for quite a while.

Perhaps a financial institution breaking ranks, but it could also be the CIA looking to test their intentions for bitcoin on idle hardware.

Looks like the GPU is going the way of Barney the dinosaur LOL !

Bring on the FPGA fools ! I never bothered with GPUs anyway !

So the age of the GPU is coming to a close? We knew it was coming, I just thought it would take longer.

Ah well... I went in hoping for a free video card, and I ended up making a pretty solid profit on top of it. I can't complain. Maybe I'll start learning about ASICs or something... or maybe I'll just enjoy the chance to play Portal 2 without feeling like I'm burning money.

I would like someone to show the calculations proving that fpa or asic mining is actually more cost effective than a decent GPU.

Indeed me too but cannot ignore how CPU is dead even if I have free electricity etc.

Price is no object here it seems so the more power a technology has the better. Eg cost of this don't matter. If it can take out gpu in speed then gpu dead etc.

The reason is that CPU is not cost feasible anymore, even with free electricity.  ;)

What we need now are some bounties posted to get FPGA technology open sourced ... links to good cards and s/ware and set-up guides ... exactly as was done for GPU back in Jan/Feb.

If the bounties are posted in BTC then the cost to the outside world becomes slightly less relevant since the the current hardware has probably paid for itself, BTCwise. It is just another demonstration that BTC is actually backed by the hardware/software and electricity going into keeping the network competitive, secure.

FPGA miner bounty, who's up for it?

I would rather a bounty be created than somebody do it on their own time/for themselves. Message me if there is one ever going..

Why is everyone getting upset over FPGAs? They're not a threat at all, and when they do become a "threat", they will be freely marketed to miners.

The best price (academic) I can get on an Altera 115K is $330. It is theorized they can do about 100mhash/s (currently only 80mhash). That comes out to $3.30 per mhash, compared to the $0.50 per mhash of a 5870. Thus it would take over six times as long to breakeven on the hardware, which increases variability (risk) by a ton. Not only that, but FPGAs have little resale value. I can easily turn around and sell my 5870/5970 for what I purchased it for.

A 1 thash/s FPGA setup would cost roughly $2,000,000 to $3,000,000 depending on if a custom board was made. Doesn't make much since, especially considering that bringing that much power to the network would raise difficulty, and decrease your profit. Might as well just pump and dump the currency itself to see a better return faster.

The ONLY advantage of an FPGA currently is its long term sustainability. Difficulty will get to a point where GPU miners can't mine for a profit after electricity, so FPGAs will have to come in. Or the huge decrease in hashing power after that point will lower difficulty and make it profitable for the GPU guys again.

Funny that the fpga conspiracy theory jumps right to the financial sector.  I know one shop that does a lot with CUDA and heard that CUDA's the reason Amazon went with Nvidia cards.

Had a consult with an engineer today and the short of it is that going out and buying the fpga's just to do mining is unlikely to be profitable.  ASICs will cost even more, but if someone really wanted to throw 7-8 figures (USD) at it, they could exceed the current hash rate of the entire network.

The best way he came up with is to do what cypherfox is doing - find someone who already has a ton of idle capacity (making acquisition of the fpgas a sunk cost), and partner up.  His connections were not private sector, so I have no leads on idle capacity at the moment.

I'll have to agree with Cheeseman. I figure GPU mining will still be profitable at least 2 months from now (provided the giant space monolith on the Moon doesn't figure mining for BTC is a fun pastime) at least profitable enough for people to break even... What we will see though is a slight settling of the difficulty rate up to a level where is worth it but just barely, increasing the time for a ROI. Then the difficulty will drop slightly for a couple of cycles, then rise again to those levels. When a situation like this is reached, that's when FPGA/ASIC mining will start being used more and more and we'll have miners slowly adopting custom boards. Then we'll have a mix of GPU/FPGA mining. This will make the difficulty levels creep up extremely slow.

The reason I think at some point the difficulty will stagnate is that, while new people will find out about bitcoin and try to join in on the mining craze, they'll see it's not worth doing it without some major hassle. So with a slight variation, the difficulty rate will be settled at some point and will only rise when new technology is introduced in the network. We will have reached the limits of our available computing technology and we'll have to wait for it to catch up.

The stagnation of the mining rate, coupled with increased popularity will mean the bitcoin's value will go up. This will invariably change the profitability rates and will get the ball rolling a little more, albeit slowly.

What I'm saying is that we'll reach a point where the profitability will be much smaller than today, and it will be linear relative to the difficulty climbs. But GPU mining will not cease to be profitable too soon, although yes, profits will probably be ridiculously small eventually. Also, if the community FPGA thing flies, custom boards will slowly start replacing GPU setups, but make no mistake, it won't be as fast as the GPU replacing the CPU.

Before anyone jumps on my head, yes, investing in BTC right now is a much more sound strategy than buying hardware and setting it up for mining. Following the general trend we've seen since a year ago, the bitcoin will only grow in popularity in time. This means more transactions, this means more circulation, more exposure and by extension more value per BTC. But mining will always have a niche and people won't stop doing it, it's a nice natural mechanism forming around the network.

Of course, everything I said here could be moot provided some higher power decides to fuck around with the network, but I don't see the governments trying to curb the Bitcoin craze while they don't see it as a danger to the status quo. At least not until next year.

[edit] Perhaps I need to give some clarification on the "next year" thing. How should I put it... You see, when animals are in the mating season they do all kinds of crazy things. Well, next year is an election year for a couple of countries I know; it's mating season for the politicians and you see, politicians tend to do all kinds of crazy things when in mating season. I'm not saying an apocalyptic crackdown is inevitable, I'm just saying the more popular Bitcoin becomes, the more it's bound to raise some eyebrows. This can also prove to be a good thing (governments attacking Bitcoin). I'd laugh my ass of once the Streisand effect goes in full gear over something like this.

That was my original contention .... now I wonder where there are scrap-yards full of discarded FPGAs??

Where are these FPGA's with "little resale value" being resold?

Looks like Moa beat me to the punch... maybe they end up at auction houses in silicon valley or there are government regulations mandating their disposal...

I could imagine going to my local auction and they suddenly have a pallet of them and no one bids on them because they have no idea what the hell they are.

Yep, could easily happen.

Wall St. and City of London have been knee-deep in them for at least the last 5 years. They use them for real-time, super-fast, algorithm trading. Sure to be some scrap coming outta there since they are always upgrading to latest to keep up with the competition ... hey why does that sound familiar?

Well they have stock exchanges all over the world...  Melbourne, Singapore and so on. I know a guy that got a free computer once from a stock broking firm that was always upgrading so its possible there's a nice guy out there that would love to us have his old FPGAs

Hmmmm ... that's got me thinking I'm gonna make some enquiries ... I knew a guy who was supplying them into the City a few years (5?) back.

With more than 50% you can just define every transaction that is not in one of your blocks is invalid, by not validating the blocks found by others. There's no need to double spend or counterfeit.

If so they have become really good at turning them on and off at the same time:
http://bitcoin.sipa.be/speed-lin-2k.png

NOW there's been a spike.  Big time.

Hrm.

Where can I find some FPGAs in the City ??? Do you mean London uk financial district eg the city ??

Located here and would be quite interested etc. thanks !

Note that the spike goes in the wrong direction when the FPGA clusters "just went online", and also compare with this historical chart (http://bitcoin.sipa.be/growth-small.png) which shows the variance on a longer time scale.

Yup, I'm aware of both of these things.  Just pointing out that the hashrate has gone from "below the difficulty" to "double the difficulty" in the past several hours.  This may or may not be significant, but it's at least eyebrow-raising.

EDIT:  It looks like it's peaked and dropped down closer to the trendline now, though.

It has seemed pretty lumpy the last couple of days ... are there any tools to see if this is outside "normal" variance ?

Miners can trivially prove their capacity by signing a challenge nonce with the private keys of the coinbase transactions in their found blocks.

Sorry, you are just completely wrong about this.  Validity of a transaction is not determined by being in a block or not.  A transaction that isn't in a block merely hasn't been accepted (yet).

Say you had 51% of the total mining capacity.

What are the odds of getting the next block, and being able to prevent transactions that you don't approve of for ~10 minutes?  51%
And the odds of getting the block after that and continuing the blockade out to ~20 minutes?  26%
Add if you need ~30 minutes?  13%
40 minutes?  7%
50 minutes?  3%
An hour?  2%
Two hours?  0.031%

AMD 7000 series round the corner in the next few months, i'm starting think the GPU's are just gonna progress far quicker and faster than than the FPGA can keep up with for cost ratio, i mean the next gen cards are likely to be close to a 1 ghash each, save your money and wait for the 7 series.

Put your soldering irons away your just going to burn yourselves.  :D

You don't get it. If you have more than 50% of the mining power you can just continue searching for a valid block even after somebody else has found one before you. This way you can make sure that you control every block. This is listed in the weaknesses on the Wiki. Yes, you can technically still add transactions, but there's not much point if they'll never get confirmed.

No, he gets it, and you are still a little behind on the curve.  Having control of 51% of the hashing power of the whole network makes it possible to successfully attack the blockchain for a short period of time.  That period of time being one 10 minute interval.  The whitepaper doesn't go into detail about the odds of success of such an attack, other than to show how it's not really possible at all at less than 50%.  Having just over 50% of the network hashing power doesn't really give you very good odds of success past one block, and an attacker intending to deny transactions into the blockchain for longer than one block has to be able to be certain that no blocks can sneak in under him, for if one gets in and the next is built on top of that before he build one to overwrite that one and one on top to secure his false one then it become exponentially more difficult for him to overwrite two blocks back.  In practice, an attacker wishing to keep this up for an extended period of time needs at least double the hashing power of the network because it's like the attacker is trying to wade up river while the honest nodes are wading down river.  And even with double the rest of the network, some blocks are going to slip in and be covered up again anyway.  At which point the attacker has to choose between trying to overwrite two blocks and then write another before a third is made by the network or simply ignore the one that got away and overwrite the last one to take the network back.

Read it again.  It doesn't say anything even remotely resembling what you are claiming it says.

Generating a valid block after the rest of the network has already moved on to the next one is very nearly pointless.  Do you understand what circumstances must be present for someone to replace the existing chain with a side branch of their own?  How many times do you think they could get away with what the network operators would consider to be a very major event before someone notices?

Did anyone see this yet?

https://github.com/progranism/Open-Source-FPGA-Bitcoin-Miner (https://github.com/progranism/Open-Source-FPGA-Bitcoin-Miner)


found it through http://www.bitcoinminer.com/post/5672391181/open-source-fpga-bitcoin-miner-announced (http://www.bitcoinminer.com/post/5672391181/open-source-fpga-bitcoin-miner-announced)

Thoughts? 50MHash/s or 80MH with proper cooling with an DE2-115 Development Kit (this is not a DE2. It has a Cyclone IV EP4CE115.)

80MH/s for a $600 piece of equipment?  No thanks.
http://www.terasic.com.tw/cgi-bin/page/archive.pl?Language=English&CategoryNo=139&No=502&PartNo=7

Good find nonetheless.

See your private message.

Thx.

Lol, actually I did already see it, since the author announced it on this very forum.

http://forum.bitcoin.org/index.php?topic=9047.0

32x bigger[8x ? didn't saw bigger chips/matrix] FPGA's, made on modern 0.045 waffer[mean working on clocks, lower(cuz arch limitation), but comparable to modern GPU's]can flush GPU's out of business :P and reinforce bitcoin :) or ruin it[cuz majority of inviduals can't afford it at 1st time].

The market for GPUs is millions of times the size of the high-end FPGA market, and gets MUCH better press.  As long as AMD doesn't do anything to make mining less efficient on their chips, we will have nothing to fear.

press ? what for ?
market ? how GPU's market created/grow ? have it anyway related with FPGA ? with or without BTC mining context ?
actually in newer AMD CPU chips[2 weeks before launch] dramatically improved sheduler/prefetcher[4-6 faster decoding/prediction/fetching of subs/micro/nano-ops/dataportion]as well as 128[finally !! hello real-quick SSE2 ! among other math thing :P]FPS, but only four per 8 int exect pipes :/ so new AMD chips could compete with iNtel SB chips very well [in case of "8x core Bulldzoder vs 4x core SB"], esp if prices was competive. and they can be - recomeded price for low-end 8x core FX chip is $210, ie same as today 6x core Phenom II 1100T CPU. and behave more interesting in scientific computation :-P
for further detail - study a Software Optimization Guide for AMD Family 15h http://support.amd.com/us/Processor_TechDocs/47414.pdf very entertaining reading :P
<--initially im wanted just to get quick highlight on such chips, but when im start reading im can't stop, until finished it completely, then said "Wow" and smoke small cup of tea to combat stress/shock[hate/intolerate intoxication of any kind. including endorphines] and then check AMD shares cost on NYSE.
IMO only RRAM breach into market can make huger impact in IT advancement !! 0.3 ns swithing time !! isn't that cool ??! for about last 12 years memory subsystem become most serious and dramatic bottlenecks[thanks, Rambus to you contribution  into memory biz R&D stalemate !!].

When a new video card comes out, hundreds of thousands of people know about it, and there are big events with reporters.  When a new FPGA comes out, twenty sweaty engineers know about it, and the press is nowhere to be seen.

you don't get it.
sometimes press/PR noise don't help, but dammage effort to create something, bring it to market and etc.
both because competing for cooperation, conflict of interests[how many ppl buy GPU's for mining after this project rise to full-power ? how many NSA/CIA/FSB crunch-farms become deprecated waste of power/iron/silicon ? how quick nowdays chiper's become deprecated too ? and other examples[im intentionaly not enumerate rest very obvious examples of utilisation of such project fruits].
i mean if you plan grow its "'from ground" its okay ? both as BTC backend or "just for fun" or both, but PR support frequently more dammage than help for such start-ups. thats outsider opinion and strictly IMO, but i seen many very amazinng IT projects, collapsed just because big-biz sharks wanted no danger for their high-margins outdated-garbage-shipping "business".
sorry if im not understand something or explain inconsistent or do it wrong way, as well as for my "engrish".

+1, exactly my thoughts

they become threat, after SHA256 become really unusably unsafe[quicker than initially expected], cuz GPU's isn't very suited for majority of other hash'es crunching :P
while FPGA's stay afloat :-)

but by then, fpga code and hw (or ready-made mining stuff based on fpga) might be "freely marketed to miners", so anyone sitting on then-unsuitable GPU can switch.

Also, I'm not sure GPUs are unsuited for calculation of the majority of other hashes. Where do you take this info? It would be a great threat to people using ASICs, though.

Also: you don't seriously believe SHA256 will become "unsafe" soonish, do you?

Even if there is some kind of shortcut found in SHA-256, this would simply require that the difficulty increase to compensate.  The only reason that we would need to migrate away from SHA-256 in short order is if a method of reversing the hash were found to be feasible, in which case SHA-256 would be broken.

The whole point is that GPUs are likely to retain a huge advantage over FPGAs because they are a larger market.

If that's true, there must be some mechanism for preventing long chain splits (or rather reorganizations) that I don't know about. I mean if I have 51% of total hashing power, I can just start building my own chain in the dark, and wait for it to get longer than the "official" chain, which it always eventually will. Then I announce my chain to the rest of the network, and clients have no choice but to accept my chain as the official record of transactions. Rinse and repeat... Right? Tell me what I'm missing here. I mean sure, people will notice very quickly that something fishy is going on, but what can they do? Nobody has the authority to pick and choose from two valid chains for the rest of the network. Clients will automatically switch to the longest chain. Right??

underdeveloped/underscaled/slow/not-mass-produced things fail anyway.
and FPGA hardly make any specific there.

What you are describing is a slightly different attack.  This is the double spend attack in a nutshell, and not a denial of service attack.  The DOS requires live participation in the network.

Still, it's not about the longest blockchain, it's about the one with the greatest proof of work.  But yes, hashing a false chain in the dark, so long as you have a least as much power as the remainder of the network, will eventually get you ahead of the true chain.  51% of the network still doesn't cut it, though.  This would only give you 2% more power than the rest of the network, meaning that you would only be able to get one block ahead every 50 blocks or so, and if the rest of the network continues to grow as it has for the past two years, your still a loser in less than a day.

hm, start investigating/counter-intel mining pools creators ? :P
bloody KGB hired Tycho to ruin BitCoin mastermind plans ? :-)
how to fix/resolve it ? insisting/force pool creators/maintainers give control/move physically pool-related infrastructure to bitcoin foundation or affilated entity[mining-dedicated sub-division ?] ???
or just scrap pools support ?

Double spending should be easy with 51% (meaning of course on average 51% of total hashing power during the attack) as well. Buy something with 10000BTC, and announce the transfer normally. Start building a blockchain in the dark with that transaction going to your own address instead. Wait until whatever you bought is delivered to you, and your dark blockchain is longer (greater proof of work if you insist), and dump your longer fake chain to the rest of the network, invalidating the original transfer. Yes, with just 1% edge, if you get a bit unlucky, you might have to wait for a long time to catch up with the official chain, but theoretically it doesn't matter if it takes days or even weeks. Practically you would want to have a bit more than 51% like 55-60% to make it easier, or maybe DOS some of the larger pools to give yourself some edge. But still, it's very much doable.

It's of course possible that I have misunderstood something.

should/can http://forum.bitcoin.org/index.php?topic=9137.0 trick/approach solve this problem ?
or complete pools decomposition/takeover really necessarily ?

I think the solution you referenced will work to resolve the issue.

There are some caveats and details to take care of, however, and it will probably take a while before someone implements it and gets a pool going.

For now, my suggestion is to let no pool go >50%. I never mine in biggest pool.

Let me toss out some rough numbers. Please correct me if I'm too far off.

You have about 300 chips doing 200 MHash/s? That's 60 GHash/s you brought online? While that probably puts you in as one of the top miners, that's not a significant cause for the recent network growth.

For someone without piles of existing fpgas sitting around idle: If purchased new, that's $250k in hardware costs bringing in $10k per week in bitcoin value. Half a year to pay off the initial investment (assuming price/difficulty growth remains steady). GPUs would break even in less than half that time.

Don't forget, that was two clusters out of an unknown larger number of them, for which the hardware was already paid for capital.  Those fpga clusters that he speaks of were paid for by other projects, and would otherwise sit idle earning nothing between major projects if not for Bitcoin.  And they are way more energy efficient at hashing than even a GPU, which is way more energy efficient at hashing than a CPU.

I should remind everyone that there has been absolutely zero evidence of any big new arrays, FPGA or otherwise, coming online recently.

Extraordinary claims require extraordinary proof.  So far, there has been zero proof, even of the ordinary kind.

I was wondering about this recently when I saw a post about a recent dramatic clime in the overall network hash rate...

Am I correct in saying that the hashrate for the network is an estimate based on how many blocks are solved versus the current difficulty? So that while it's statistically likely to be fairly steady there can be lucky and unlucky days even for the whole network combined?

Just editing to add that of course it can also be a lot of new GPU miners coming on-line in combination with the above when only looking over a short period.

I would imagine that whoever owns such hardware has some real use for it which he expects to profit from, despite the large initial costs. Bitcoin-mining with these clusters is profitable right now, but since such hardware power can easily be used (and sold) for other profitable projects in the industry, the threshold for them is most probably not electricity cost, but the next most profitable project the hardware could be used for instead.

Don't forget, that with all the home-brew GPU mining rigs only electricity cost is the limit because their computing power can not easily be sold for anything else. If difficulty jumps up, the profitability shrinks for everyone - even those with the most powerful hardware.

Another factor is, that we have a global competition for energy prices and I am pretty sure someone here has access to cheaper electricity than whoever operates those clusters.

Last but not least, GPUs are getting faster and more efficient as well. Due to the much larger market and the very low hardware costs compared to FPGAs or ASICs, the advantage of the latter two shrinks continuously and the cost to upgrade them to the next better hardware generation is vastly greater.

FPGAs and ASICs are by far not the most imminent danger for Bitcoin so I think we should all just relax a bit, congratulate cypherf0x for his current advantage and upgrade our rigs with Southern Islands when they are available.

Everything's going to be just fine  ;)

The price of the cards I use for GPU mining initially cost me $270 per.  Comparable cards cannot be found on newegg.com, and the "out of stock" price listed is >$500 per card.  AMD and NVIDA cards are just not competing.  My bet is that ATI knows exactly why their cards are selling so well and would be willing to create low cost ASIC, FPGA, or GPU hardware specifically targeted to mining.  If I were NVIDA or AMD, I would be searching for ways to cater to this market as well.  

If you are really concerned about a threat to FPGA centralization of computing power, start an email campaign to NVIDA and AMD asking them to create hardware specific cards to meet our mining needs.  All that is needed is a bit more competition.

EDIT: Cleaned up awkward wording.

Right now that's no more than 10 000 AMD GPU mining,   Does that really matter for AMD ?

I'm sure some product line manager at AMD has noticed the strange selling out of 6990s and 5970s, and if they have even an ounce of grey matter they've deduced it's the result of Bitcoin. But to support the development of a custom ASIC by one of the big boys, the card sales volume needs to be truly astronomical. I don't see Bitcoin really changing the course of NVIDIA or AMD until there can be provable demand for millions of such custom cards.

In the mean time, there is definitely a niche market for a custom Bitcoin ASIC, if someone can get the job done profitably.

This is not correct.  You have to remember that the chain with the highest difficulty is the one that the network accepts.  If you control >50% of the mining power, you do not have to build on anyone else blocks, you only have to build on top of the ones you produce.  Over time, your chain pull away from the chain that the rest of the network is producing and your chain will be *the* bitcoin chain.  This attack would afford you the opportunity to double spend, or block transactions that spend out of certain accounts, or block transactions altogether...and probably many other kinds of mischief.

What you are missing about this attack is that you have to start your chain the minute you give someone coins. You can't precompute a bunch of blocks because to double spend, you have to first spend.

So you send someone money and now you're working on a chain. If they wait a few confirms (the paper describes in detail how quickly the efficacy of this attack drops off with number of confirms) then you have to have beaten the network for X confirms. That is, your hidden blockchain fork has to be at least X+1 blocks long when they release whatever they are releasing to you.

The chance of that happening increases as your power goes up in relation to the network, it's true. But you are working on a blockchain completely independently of the network. You want to do a lot more than equal ("51%") the network - you want to at least have 1.5x the power. You need to be able to find blocks faster than the entire rest of the network, and you need to beat variance and do it consistently for long enough for your transaction to be confirmed "enough" for whatever you're doing.

And if the honest network happens to get luckier than you and get ahead early, you're screwed. Catching up gets exponentially less likely with every block they're ahead. If they get two blocks ahead, well, now you've basically wasted a bunch of processing power finding all those blocks that you COULD have used to generate blocks instead. Maybe your scam would have paid much much more than that, but if you fail you're out the bitcoins (i.e. the extremely large amount of bitcoins you're trying not to spend, getting spent) and the generation profits.

I think you vastly underestimate the number of non-bitcoin users of these cards. At best, bitcoin would be lumped in with the other distributed computing projects that can use GPUs. And don't forget the armies of actual gamers who purchase these cards. Bitcoin, although large, is very much just a drop in the bucket.

Also, you will not see a bitcoin ASIC. Complex ASICs required highly skilled (read: expensive) engineers and even more expensive processes to put them into place. When there is already a very good solution to the problem (GPUs) and an enormously volatile and unpredictable exchange rate, you're not going to see such an investment. Not to mention that investors in such a project would almost certainly demand immediate cashing out of mining proceeds to pay off their investment, which would put significant downward pressure on the exchange rate. ASIC mining is a pipe dream.

I hate to disappoint you...

Unless you're referring to those had to read your vague half sentence, I don't think you've disappointed anyone.

Are you suggesting that you know of someone putting millions of dollars into developing a custom ASIC for mining?

You mean other than ArtForz (http://veritas.maximilianeum.ch/bitcoin/irc/logs/2010/12/20#l2461)?

ArtForz designed and ordered himself some ASICs (het put two rounds of sha256 unrolled plus H==0 logic on them, if I got that right, and I don't think one would call 2xsha456 "complex") around January and already received the first 2 batches (100+100) (this might even be old info by now, I think he might've been planning for 1000, not sure).

Some weeks ago the first 96 where already online. They each do 200MHash/s using only 8W. He controls, I think, 8 of them on a small board with an fpga to feed the work and receive the solutions, which he plugs into a backplane of some sort. All 96 fit in a 4U casing.

So your assertion "you will not see Bitcoin ASIC" is wrong, it's already here, ask him on #bitcoin-dev.

And yes: the guy is highly skilled. But "expensive process to put them in place", I don't know: he said he used a modified toaster to do contact soldering ;)

Question is how much did he pay for the whole process? Tens of thousands of dollars?

200 Mhash per chip with just 8 W is amazing though!

IIRC, ARtForz mentioned at one point that his cluster had crossed the $200,000 mark back in the Fall of 2010.  So we aren't talking about your average geek throwing this together in his garage.  ArtForz has resources.

God daaaamn!  :o

Yeah, that's one serious bitcoin investment there. It makes my several thousand in mining hardware seem like peanuts. Well, good for him! I do hope he uses his bitcoin wealth to further the bitcoin economy and to start new and interesting ventures. I'd also hope he'd sell me one of his magical ASICS but hmm doubt that will happen :)

P.S ArtForz, I can pay you in bitcoins.

Bitcoins have failsave  to resist case of "over 50% hash power"  failed in one hands.  I am not remember when, but think satoshi mentioned that in some update notes.

ArtForz is to mining as Satoshi Nakamoto is to software.

The guy is a legend, does he really exist? :D

Next thing you know there'll be a rumour that he whipped a fab. room in his backyard from old tractor parts and TV sets to pump out ASICs!

Hahahaha...:) I'd pay to see his mining operation in action. Geek heaven that's for sure!

From the sounds of things these sound like structured ASICs - somewhere between an FPGA and a normal ASIC in terms of performance but much cheaper to produce.  I could be wrong, but that's what it sounded like to me.

That's what he said. Works sorta like fpga with pre-fabbed logic gates. Then, through some processing magic, connections are permanently made (something about metal). Well, you know better than I do how that works and what it is exactly, but I remember him describing them along these lines, so you're likely right.

I'd be willing to finance engineers to build a solution then I will resell it,
but first i need to know what to tell them?

Can we use this board?
http://www.knjn.com/FPGA-PCIe.html

Sorry for being vague. Yes, it will be announced soon that someone is putting a great deal of capital into custom mining hardware. Getting a few ASICs produced using a shared wafer process is not particularly expensive. This is larger scale.

At what price does an ASIC batch start at ?

(rough estimate please)

http://www.edaboard.com/thread115517.htm (http://www.edaboard.com/thread115517.htm)

That discussion is a couple of years old.

Also see this PDF, also a bit old:  http://www-ee.ccny.cuny.edu/www/web/xchen/StructuredASIC.pdf (http://www-ee.ccny.cuny.edu/www/web/xchen/StructuredASIC.pdf)

You can get a design into a shared wafer for as little as $25,000. In some countries there are academic fabs that make chips for student projects, etc.. and these can be had for much less, but you need the hookups. You also pay a lot less for older processes.

Is it possible to have an FPGA or ASIC board that runs independantly from a computer but has flash memory onboard,built in Wi-fi or Ethernet,and has it's own power supply as I wish to bypass the computer competely and save loads of energy while mining lots more.How much would these boards cost and for what performance/power use can I expect from them? Can yiu ship to UK? Do you have a shop?

You guys might want to take a look at a new thread I started regarding a design that we're looking to sell. We think it is a very competitive offering but we don't have the time or energy to take this to market. Check out the highlights here and let me know what you think of the performance, power and cost numbers.

https://bitcointalk.org/index.php?topic=44391.0

FWIW I haven't been able to get a quote under $70,000 out of MOSIS, which handles the majority of academic chip fabbing.

Usually "the hookups" equals "fab has been hiring our lab's graduate students for a decade now and wants to keep the supply steady".  In which case there's actually more remuneration going on above and beyond dollars changing hands.

We've obviously looked at this a great deal at LargeCoin. The cost of fabbing a profitable Bitcoin ASIC is $500,000. This is the cost when you use a structure ASIC solution, and there isn't any way to get it below $500,000 - believe me, we have tried.

Structured ASIC processes definitely remove a lot of the complexity of ASIC fabrication versus standard cell or full custom processes. However, you are still dealing with the fabrication of an immensely complex integrated circuit with millions of tiny little wires in it. Getting the design right so that the first batch works means spending lots of money on sophisticated software and engineering talent - and that's just not cheap.

If you want to run off a few chips in an academic process, by all means go for it. That's a great hobby level activity. However, we found through our extensive analysis that nothing short of a structured ASIC process can give you the scale and volume that you need to generate enough money from your efforts to be well compensated for the investment.

Are there any news regarding the progress of your product?

Hard at work is all I can report.

I see... Good news then ;) Good luck!

Thanks. Sorry we've been so quiet. I'm glad to see that Bitcoin has turned the corner after the summers' scathing press coverage etc. The mainstream seems to be picking up on the actual utility of it, which is nice.

I'm actually curious about FPGA's is there like a list of component/design files floating around that anyone knows of?  I'd like to read more into it and maybe construct my own since I have access to a pcb mill and soldering tools.

Currently, the Xilinx Spartan 6 LX-150 is the most common FPGA used for bitcoin mining.  GITHUB has an open source FPGA Miner for this chip, and ztex.de does as well.

Unfortunately, you won't be making a PCB for the LX-150 on a mill.  It's a 484 pin BGA.  It pretty much demands a MINIMUM 4 layer PCB because of the split power needs (3.3V and 1.2V) and ground.

Enigma

I have one of these for sale if anyone is interested.  Brand new, never used.  PM if interested.  I just never got around to that project.

Thanks for the heads up, I was just curious as to the process and what not my background is Comp Sci, not so much Comp/Electrical Eng so I don't know much about FPGA's.

I hope that at some point in 2012, I can come back and reread this post and understand what all of these words/letters/numbers mean, haha.

Would a glossary help?

Xilinx is a chip maker.  Spartan 6 is a line of FPGA chips.  LX-150 is a model in that line.

FPGA means Field Programmable Gate Array.  It is a chip with a large number of generic logic gates and programmable interconnections.  This means that it can be configured and reconfigured to do different tasks as needed.  Think of it as a CPU that can be optimized for a particular job.  That's a bit of a simplification, but I think it gives you the idea. 

Github is a public source code repository.  Someone has posted source code that can be loaded into a LX-150 FPGA chip to configure it as a miner.  ztex.de also hosts similar software.

PCB is a printed circuit board, the flat board that electronic components are usually attached to.  Most of them are green.  The board itself is usually a composite insulator material with sheets of copper bonded on the flat sides and then partially removed to leave different connections and circuits.  Commercial boards are etched with acid, but a variety of methods are used for prototyping and home production.  Milling when a rotary tool is used to mechanically cut out portions of the copper layer, usually using a computer controlled machine (a CNC mill, but that's just another acronym to define).  Connections between the sides are done by drilling the board and connecting the copper traces on both sides.  In commercial boards, interconnects are electroplated chemically using processes that aren't generally possible at home, so homemade boards usually use pieces of wire soldered on both sides.

Boards with more than two layers can be made by stacking single sided boards onto a dual layer board so that you end up with alternating layers of copper and insulator.  In commercial boards, the interconnect holes are plated as each layer is added, so it is possible to connect any layer to any other layer.  This is nearly impossible on homemade boards, which is unfortunate because you really need those internal connections (called "blind vias") when working with BGA chips.

BGA means Ball Grid Array.  It means that the chip's pins are all on the bottom, and not along the sides.  The chip is manufactured with a square grid of solder balls on the bottom of the chip, and it is placed onto a PCB with a matching square grid of copper pads and the assembly is heated in an oven until the solder melts and connects the board to the chip.  Because the grid has hundreds of balls (484 in this case) with very tight spacing, there is no room for all of the traces to leave the area on the surface, so they must be connected to internal layers in the board stack and routed out.  Even doing it with 4 layers is rough, 6 is way better.

There are several successful 2-layer miner designs now, with no problem getting 15W+ into the part.
It takes 6-7 mil trace/space, ~12 mil drills and a good soldermask - so it is still definitely not a "CNC-at-home" PCB.

-rph

I suppose it helps a lot that bitcoin mining doesn't require much data, so only a few IO pins are needed.  Do you know if these 2 layer boards have any thermal problems?  I know that some BGAs can't dissipate enough heat out the top even with heatsinks.