Bitcoin Forum

I'm starting this thread to find out how many Newbies are here to look into FPGA mining or some other use of FPGA but using the current work done in Bitcoin as a launching platform.

+1 for me. I created this account just to research an FPGA miner I could also use for electronic hobby projects.

A.J.

FPGA's are amazing, but can they build specialized enough hardware that beats the monstruous processing power of GPU's?

Do they live on e.g. a PCI card that you plug into a normal computer, or is it something more painful to use?

USB connectivity so very easy to implement. Programming the FPGA seems to require JTAG on some boards which is a bit confusing. Why not use USB for all communication?

A.J.

@bluetrepidation I think JTAG is in the pinout of most (all?) microprocessors and FPGA's. JTAG is a debug port standard, different from USB which is general purpose. You can just buy a JTAG-USB adapter and live happily thereafter, though :) It wouldn't make sense to force one of these inside every PCB or IC; it would just take up space.

@nostromo429 there are FPGA kits with PCI and PCI-E connections, but these are only used when very large amounts of data need to flow between the FPGA and the PC. I think for Bitcoin mining, the USB speed would be plenty.

same here :)

does anyone know where to buy samples of
both pcb's
Ultra-Low-Cost DIY FPGA Miner - 175MH/s @ $1/MH
https://bitcointalk.org/index.php?topic=44891.0

I don't have enought post to ask question on above forum

Ive been researching fpga's due to the high cost of electricity... But the price to performance has been high...

+1 here too... $1/MH is not bad at all considering how much you would save on powering them...

+1 here... have a few FPGAs already and am just about to start soldering a new PS for them... just wish the FTDI chips support ARM and I could ditch the using PC's all together....

I am willing to buy a BFL unit if they are for real.

Yeah... I keep seeing people showing theirs off, but I have yet to see someone selling a low-cost package.

Well FPGAs are very interesting im interested in any cheap affordable mining. If only FPGAs where cheaper i would invest in hundreds.

what is the average hash rate of these FPGA boards and at what wattage?

what is average price of an assembled board?

can once save a good deal of money if they knew how to solder? 

1. 350-380 MH/s per dual-chip Spartan6 board at ~20W.
2. $560
3. Not really a good deal because you'll lose lots of time for manufacturing the PCB and collecting all the components.
Also you should remember that soldering BGA chips is not a task for newbies.

If we plug the figures from
http://cablesaurus.com/index.php?main_page=product_info&cPath=10&products_id=56
into
http://bitcoinx.com/profit/index.php

I get a minimum almost 6 year break even for UK prices and that's before the drop to 25 blocks this year (current price $5.7, 0.19elec, power 25w estimate beyond the 20w minimum, 250mhash). After the rate drop the calculator just says >10000, can't get it to work.

It's not economic yet but it will probably be the future as the cost comes down.

I also joined for FPGAs :) Looks like there is a lot off interest

Have you guys been looking at the Butterfly labs page in the mining forum? If they actually put out a product I will certainly be buying some.

I joined because I bought a ztex FPGA and it's the future of BTC. Now I mine 1 BTC per day week for 8W, silent and without ecological regret! GPU miners that don't recognise FPGA mining are just bitter. They need one PC per 3-6 cards while FPGA is controlled via USB so you can have 127 FPGA's on one PC!!! Do the math on that and the electricity bill...

FPGAs are not good for people like me with low electricity.

They are best for people paying like 0.3 USD / kilowatt hour etc.

ATM, I think I will stick to Radeons, thank you very much !

If BTC dies then who will be buying your useless PCBs ?

Heresy! Bitcoin is special and magical and great and makes its own sauce. If you even allow for the possibility of anything else being true you are in the wrong place, go away bad person!

How do you like your ztex? I general hear positive comments about them from forum members.

The good thing about FPGA's is their very low power consumption. However, their $/Mhash ratio is also very low. It is really hard to justify their purchase from an economic standpoint unless your electricity rate is very high. Generally speaking, a GPU will net a higher profit for your investment (even when taking into account electricity costs).

An FPGA will currently run you ~$500 for ~350Mhash/s. For that same $500 you can buy 3x 5870 GPUs providing 1200Mhash/s. Yes, the GPUs will be using a lot more electricity. But they will also be generating a lot more profit. And I haven't even brought up the issue of the additional PC you will need to control the FPGA ... but let's say that equates to the motherboard/PSU/RAM you would need to host the GPUs.

FPGA mining boards of today are much like today's gasoline-electric hybrid cars. While they are a very interesting and efficient alternative to traditional GPUs (gasoline cars), their high cost makes them economically unviable in most cases. Unless, again, your electricity costs are sky-high.

I think FPGAs are the future of mining, but that future is not quite here yet. There will always be the early adopters paving the way for them to become mainstream.

I know for a fact you don't get 1 BTC per day on 8W.
http://www.alloscomp.com/bitcoin/calculator.php

More like 0.16 BTC per day.

While FPGA may be the future GPU aren't dead yet.  To replace the 8.85 BTC I earn per day I would need ~55 ztex boards which even w/ discounted pricing runs ~$16,250

Exactly. Currently it is difficult, at best, to make a case for FPGAs in terms of economics. But that doesn't necessarily make them a bad choice for some. There are reasons other than economics ...

DeathAndTaxes: Ah, sorry the 1 BTC is per week.

Epoch: The ztex is great because it's silent, I don't care about getting rich, all currencies make you poor at heart! ;) Also you only need 1 PC per 127 FPGA's = alot better than GPU mining.

enigm4: We'll see which PCB's become worthless first. My bet is that peakoil is going to make your PC's and GPU's completely obsolete. Personally I run an Atom machine at 13W for my everything PC (24/7 server AND development station) it's completely silent with a SSD.

Why would peak oil significantly affect the price of electricity given 99%+ of it isn't produced by oil?

I can't imagine many horrors worse than compiling a major project on an Atom processor.   Well I take that back yes I can.  database query optimization on an atom processor.

What do you think will "try" to replace oil? What do you think is causing the current economical crisis? The fundamentals have been choking our civilization since the seventies but we hid it in the currencies/"exploitation of natural resources" (pollution) and now we see the unwinding of our unsustainable lifestyle.

That's why I'm betting on BTC, because anything other than old civilization fiat is good.

My rough calculus of intrest rate on FPGA is ~ ($ per bitcoin) * 10 = 65% right now.

And that's enough, I don't need another "not work" work trying to get rich to buy more things I don't need. I wan't stability and preservation = I need small scale diversification.

I also have 10+ years of Java development, and I have built my own web server and database abstraction toolkit, if you need large projects (all my projects compile in < 1 sec on the atom) and don't know how to optimize SQL post launch you are probably building it wrong way in the first place.

But you are right Atom is not for learning stuff the hard way, Atom is for when you master your trade and build your own tools. But you would be surprised of the speed of this little machine! I watch 720p flash video, no problem!

Care to share that calculus because the interest? ROI% FPGA is nowhere near 65% annually.

I = 300 EUR
R = 52 BTC * 6,5 USD = 338 USD = 264 EUR

Actually it's 88%

LOLZ.

Ok, ROI is -12%, but I'm happy with what ever you would call my 88% above, I call it interest rate because I would get like 1% for those 300 EUR in the bank (-2% with inflation) now I get 88% (86% minus inflation).

Instead of laughing at me you could share your ROI so we can have a constructive dialogue about this?

I'm taking your silence as a confirmation that you wasted a lot of money on those GPU rigs, and you keep wasting money on electricity every day. You can't mine on solar and wind and therefor you're part of what I call the old civilization. Tick Tack the counter is ticking.

Sure.  Your right.  Somehow I lost money on my rigs and then daily I compound the problem by losing even more more despite the obvious solution to simply hit the off switch.

11 GH = ~8.85 BTC daily.  5 rigs @ 870W each = 4.35KW   4.35KW *24 * $0.092 per kWh = $9.60 daily
8.85 BTC daily & $9.60 electrical cost daily.
$9.60 / 8.85 = $1.08 electrical cost per BTC.

Since we all know today the price of BTC was below $1.08 at Mt. Gox I lost money yet again today.

Still maybe I will be "luckier" tomorrow?

What did those rigs cost? ROI?

In which country do you live?

Sigh.

Rupy, the thing is that those GPU rigs cost a lot less than the $15k it would take to create an 11Ghash/s farm out of ztex fpga's. ROI? D&T has been around for a long time; his rigs would have long since been paid for by mining profits.

<D&T, sorry for the hijack.>

Yeah, I have FPGA fever... :P

Actually I'm considering building a "heater" rig to have running in my summer house during winter to keep the house dry now.

Thx!

I am interested in replacing my gpus with fpga hashing... hopefully I can increase my total hashing amount when I do switch over.

Hello

I am not that long busy with mining but those FPGA sound great if you want to start or need to invest in hardware if you have nothing great in your PC.
There are FPGA who do 300Mhash/s and with the curent difficulty you get about 44dollar a month when the excange is 6 dollar.
Those will use about 20Watt so when looking at my HD5670 with 61Watt and a 50% cpu i would do best to get me a FPGA.
But.....what to do when BTC is over?
Also i know nothing about those FPGA programmming and to create a file/bios for it you need a license who is not cheap and that way you have to hope people will build it for you and that is my biggest wurry.

People are already working on this.

http://bitcoinfpga.com/

This is a good starting point. Check out the links at the bottom.

Thanks for the link.

I did stumble in that topic before and found the 6500 who i am interested in. https://bitcointalk.org/index.php?topic=40058 (https://bitcointalk.org/index.php?topic=40058)
They are also willing to make new software for it but i find id scary that when they stop i have a useless FPGA or at least one with old software or with bugs.

Also the 580 dollar is a lot for 7.33 BTC a month max.
When taking 580/44=13.1 month before you have the money out when calculating with 6 dollar a BTC.
Also i don`t know how the difficulty will be in those 13 months.
Maybe you get less then 44 dollar a month and it will take you a lot longer.

Is there a list how the difficulty will be a few months from now and how long it will take to get a BTC?

There are many online BTC calculators around. Try this one: http://www.alloscomp.com/bitcoin/calculator.php (http://www.alloscomp.com/bitcoin/calculator.php)

There is no list of what future difficulty will be. No one knows that, and there is no way to predict. But you can run a few scenarios with the BTC calculator to see what effect it will have.

For a historic record of how difficulty has been evolving (for the past 15 months), check out the graphs here: http://bitcoinx.com/charts/ (http://bitcoinx.com/charts/)

The above site will also tell you what the estimated NEXT difficulty change will be (currently estimated to go up 7% in about 10 days from now).

I too like the look of the BFL unit mainly because of Mhash/$ this makes payback assuming constant difficulty and my power cost with 6$ per BTC 4.63 months, at 3$ per BTC it is 9.7 months. I realize that difficulty is dynamic and lower prices cause a drop in GPU because of profitibility and higher prices will undoubtedly push payback beyond 5 months. That having been said Since my DVR runs an ATOM and adding a different card for a GPU is impossible adding a power supply, MB, Ram, case and card I am sure it will get clost to 600$ (price of BFL unit) since I would not have to expense these things I can make a reasonable profit of 90% per BTC with my power and at 6 it is more then 95%. I am not saying the GPU folks are wrong. If you have a new enough computer it may cost you 0$ for one card (included) and a couple hundred for each additional card up to your third if you are lucky. My hardware is unprofitable or a little old for upgrades so building a rig or adding FPGA is about the same to me cost wise, but cheaper power wise. Again at current prices GPU and FPGA will pay back GPU likely faster, long term FPGA will have more profit per BTC.

Your numbers are a bit off. At 832Mhash/s and a difficulty of 1250757 (current) and a BTC price of $6USD/BTC, you will generate exactly $4.01USD per day. At $0.10/kWh, electricity costs will eat $0.12USD per day, making your profit $3.89. That makes payback at current levels in 154 days (5.1 months).

In 9 days when difficulty increases by 10%, you will generate $3.49 per day, for a payback of 172 days (5.7 months).

Keep in mind that the above numbers are overly optimistic; the calculated payback periods assume difficulty and BTC price will remain relatively constant. More realistic, however, is that if the BTC price stays at the $6-$7 level, you can expect difficulty to rise further over the next few months. It is overly optimistic to expect/ssume difficulty will remain constant.

Back in September 2011 when BTC prices were hovering in the $5-$8 range (as they are now), difficulty was 1750000. That's 40% higher than today. It may be wise to assume that if today's BTC price holds for any length of time, difficulty is likely to reach 1750000 again.

And in that case, any payback period will be extended by 40%. So instead of 5-6 months, you are looking at 7-9 months. And that's *if* BTC price stays at $6/BTC during that period.

I've had this in my bookmarks for quite some time.

https://bitcointalk.org/index.php?topic=44891.0

Have never gotten around to reading it though.