Topic: USB 300MH/S ASICMiner Value Question (Read 8321 times)

OK, its up now at https://github.com/kramble/DE0-Nano-BitCoin-Miner/tree/master/Mojo_LX9

The README should be pretty explanatory. I was unable to build the six hasher variant as PAR did not complete (overnight). It ought to fit (my homebrew Xilinx_LX9 is essentially the same code), so I need to investigate FPGA Editor to see if I can tweak the placement. So the current version is configured for three hashers (config is near the top of fpgaminer_top.v) and 50MHz. I'm using a PLL so device frequency is also configurable (if you get it working at 50MHz, try running it faster, best to use 10MHz increments to avoid weird divider ratios).

I won't be able to test it myself as I don't have a Mojo board, but I will do a port to run it on my homebrew LX9 later today (ie keeping the mojo serial code). The python mining code does work though, just be sure to set the correct COM port at the top of the file. For python neophytes, beware changing any whitespace, as space is syntax in python  

BTW This is mostly fpgaminer, makomk and technohog's code, so thanks are due to them. I just tweaked it a bit.


Yes, but only at a few MHash/sec ... dust! You'll probably never reach the pool's minimum payout at that rate!

A good job kramble your port to LX9 for other developer boards. Can you create a new folder in your git project for Mojo? with a base that we can contribute.

The first task we could do is to use the usb port and avr_interface library to communicate, how do you see?

NOTE: a small board (Mojo) that can serve to mine, right?

To chime in on the original question. The value if you're adding to an existing mining operation (or attempting to build a new one) is negative. If they could be purchased for 0.5 btc that would be much much better, but still not 'good'.

If you're buying them to replace existing gpus that are currently hashing, then they do actually have positive value in the sense of them reducing your electric consumption. Assuming you replaced some gh/s you'd hit roi in only a few months just on the power savings.

However, ~2btc each they'll take almost 2 years to earn back the btc you'd spend on them... and that's a best case scenario.

Ye that's exactly my point also. its going to go up if its going to be successful so any questions about value of equipment are only relative to how long you are willing to invest in BTC for.  The risk is the complete collapse of it. Also people who have the equipment to sell want to make their profits now because they feel like their previous risks should have paid off by now are unsure of the future like everyone else is.

It shouldn't be linearly rising. This is going to be a very complex process including many other factors. You will see it rising with the next financial crashes, considering it is a safer bet than some other notions. Like it has happened before.

Every crash left it  higher and higher. And financial crashes always come. That is a given.

WTF are you talking about? I am talking about the BTC price and you go off on a tangent about the rate of coin mining.

The problem is that BTC price is not rising in relation to difficulty increases as it should be.

The US gov has gone out of it's way to make it harder to get $USD into BTC exchanges, and there are periodic rapid BTC sell offs driving the price back down, when it should be rising.

I think that they are an investment.

BTC cant stay at ~100$ forever if its going to be "successful".
With only 21 million coins, market liquidity won't be possible with only $2.1billion money stock.

The good thing is that Bitcoins ROI time is shorter than traditional investment methods.
Its still a gamble, and with all big rewards, comes big risks - just as it has been like at every step of Bitcoins existence.

OK, I've created an account on the Mojo board to give some support on this as its rather OT here. Discussion on bitcointalk probably belongs on https://bitcointalksearch.org/topic/official-open-source-fpga-bitcoin-miner-last-update-april-14th-2013-9047

Mark

I would hope that anyone with such a board has a least the commonsense to realize that FPGA generally don't fair well  if you shove 5v up the tootsie ;-)
Personally I aim for Virtex5  Virtex6 of Spartan6 , but in about 2 more difficulty changes, they will also have the same ROI as the 300MH/S usb whit-less wonder bait.

The sad thing is that in the distant past...... even the XC6SLX9 would have provided a decent income.

I floated the idea on the Mojo board forums (uses Spartan-6 XC6SLX9)
http://forum.embeddedmicro.com/viewtopic.php?f=3&t=68

There's interest, but at the moment, it seems as though the XC6SLX9 just doesnt have enough resources for the full design.
This was referencing https://github.com/progranism/Open-Source-FPGA-Bitcoin-Miner

That being said, some cleanup could be done around the serial comm design code included the OSFBM, and the serial comm for the mojo... but I doubt that alone would be enough to get the design small enough.

... I'd love for someone to prove me wrong though and get a working miner on it  

Edit : lol hadn't looked at https://github.com/kramble/DE0-Nano-BitCoin-Miner/tree/master/Xilinx_LX9_hashers_6 yet...

You can see a basic value comparison here:
https://bitcointalksearch.org/topic/mining-equipment-manufacturers-209750

Yah, but be careful with voltage levels. Connecting to a full voltage RS232 is going to fry the fpga instantly. You'll need level translation (even for a 5V TTL  port). The serial protocol is customised too (and fixed at 4800 baud to work with a minimalist opto-isolator circuit), so its not just plug and play here. Then again if you're buying an fpga development/educational board you should expect to have to put some effort in

Oh, and to repeat, its only 5MHash/sec, or 2.34MHash/sec or whatever. Just a toy which will mine you some dust only.

[Edit] Clarifcation, by "you" I mean "you dear reader", not "you razorfishl" as you're clearly far more knowledgeable than me!!

there are a few pins spare....... Just tag on a serial port... 5 minute job....

Yup, 30 bucks. I honestly don't think the Block Eruptor USB are that awesome...

I though about building devices like these, too. With BFL chips

The Pi is better value than the Block Erupter

the pi, at least, can be used for many, many other things. everyone should have one

Then add the cost of the Raspberry Pi on top of the 2.6BTC before you work out the ROI

Smart people consider whether or not their computer is running anyway

For me it would make no difference, but the deal is shitty already without that factored in. That is why you can use a Raspberry pi to power FPGAs and similar boards: It doesn't matter, because the RPI doesnt use much power!

If your miner makes less than a dollar a day and you computer uses more than a dollar a day in electricity what then.

Smart people count the power cost of the PC to run their mining.

the 5 MH/s figure i just pulled out of thin air to illustrate even if ur over optimistic, this is still a shitty deal.

Using the usb code isn't difficult ... but don't expect 5Mhz
At 50Mhz you will drain more than the 500mA of an usb port.

Loop are rolled : 6 hashers, so do your math :

At 50Mhz : 6 / 128 * 50 = 2.34MH/s

128 parce 64 round de sha * 2

Has anyone actually tried that code on the http://www.em.avnet.com/en-us/design/drc/Pages/Xilinx-Spartan-6-FPGA-LX9-MicroBoard.aspx ? It would be interesting to know if it works. But DON'T go buying one for mining. Pathetic 5 MHash/second is mining dust. Just worthless.

[Edit] (Answering myself) No it won't work as it uses RS232 port comms. Someone would have to add the xilinx usb code from fpgaminer's github.

Yep, they'll never be profitable, but once the price comes down, they'll be perfect for educational and advertising purposes to promote bitcoin.

The code is already written : https://github.com/kramble/DE0-Nano-BitCoin-Miner/tree/master/Xilinx_LX9_hashers_6

Block Erupter doesn't even come close to breakeven with free electricity, free hosting computer, free internet and all the rest. Just don't even consider one at 2..0 BTC. But ASCIMINER will continue to sell them for as long as there are gullible punters that can't do the math on the ROI. Only when the sales dry up will they drop the price (it can't be costing them much more than $25 to produce).

So you are including the cost of having a computer running? Everyone mining has to leave the computer running. Why factor that into the cost if the point is to compare the value of mining options?

The reason I ask is that, from what I've seen using calculators that factor in rising difficulty, these things don't break even whether you factor in the cost of a PC to power them or not.

You need only one computer to run one or several Block Erupters, think about it.

My math-fu may be failing me, but why would 1 Block Erupter not break even, but multiple would? The cost, hashrate, and power consumption all rise linearly as you add more USB sticks, no?

I don't think a single unit will ever break even, multiple should. The single unit will barely make $1/day, it won't talk many doublings of difficulty to drop that revenue down so it won't cover the electricity cost of the computer it's plugged into, which all of a sudden has to stay on 24 x 7 for the miner.

if it can be mined merged with bitcoin, then most probably yes.
I was more talking about "future" applications of proof of work, based on sha256, that if the miners will be well distributed, no problem using p-o-w more frequently

Can it be used for other coins such as PPCoin at the moment?

It's true it would take awhile to break even, that's why they will come down in price (well they should) otherwise no one will buy them after the difficulty rises, this needs to be known a lot more before people start buying loads.
They can be profitable, of course. I'm thinking of buying one right now just to test it and just to say I had it.

I think mining with this hw at this purchase price is more a statement then a business plan.
the price is ~2yrs hackerspace membership and that's ok for a pet project that might bring in some reward (not 100% loss of initial costs)
but it will not be profitable (not mine the btc back)

it will be useful though for all other kinds of crypto works in the future. and thus "priceless"
like the raspberry pi helped to spread cheap computers this will bring mining to 'every household'

With the current diff increase, these things mine about 0.01BTC per day, so it would pay for itself in only 260days assuming the difficulty doesn't go up in that time! In reality it will probably take years to generate 2.6BTC

Haha... yeah, my "suggestion" is that they give them to me for virtually nothing... lol

I'm seeing Buy It Now prices on eBay that were the starting bid only a week ago, and starting bids in the $320-$350 range, if the price continues to come down the economics certainly change.

ASICminer is one of the few who delivered something that actually is in stock (not a Pr-order) and works right away out of the box, but I think that these ASICs in general are way over priced, regardless how much money you will make the first month (on This deff.factor, and BTC price), because when I invest my money I look at the long-term aspect.

A blockErupter @ 13 GH/s for 60+ BTC is really over priced this is ~ 7500 $ , my suggestion is a max of 20 BTC .
The USB miner @ 300MH/s for 2.5 BTC as well is really over priced ~ 315 $, my suggestion is a max of 0.8 BTC  .



The reason why I stopped caring about ASICs and mining BTC in general is for this exact reason, as a small miner and a father of 4 kids, there is no way that I can afford to buy these ASICs, I know that these chips are the future for BTC mining and it is really a good thing for the network security, but the fact that there is a VIP club or let me put it in other words, a club of people with a bag of money investing in these ASICs, which means that few people controlling a big part of the network, just because these prices doesn't allow all of us to participate and get the chance to be a part of it.

these company’s are making a huge profit out of you guys. the first batch of blockErupter went for around 19xx BTC this is more than 240,000 $ for 50 block , come on people open your eyes.

If just BFL had a real working chips and did deliver on time, then it would make sense to order some ASICs at the prices they are offering .

I agree which is why I sold mine for $550 on eBay local pickup only . then mutually canceled the transaction after I met the buyer in person! hah! He paid cash.

Granted there are Ebay and Paypal fees, but shipping costs should be paid for buy the buyer, and the BTC buying fee should be pretty minimal if you use the right source.  Also, a non-paying bidder just means a waste of time, as there would be no fees involved when ebay refunds the final value.

Overall though you will only see any profit in these units if you buy them in one of the group buys and sell them on ebay in the next few weeks.  Let's do the math.  Using https://bitclockers.com/calc, at current difficulty, 1200 Mh/s will return .0539 BTC per day.  1 unit can do 300 Mh/s, which would equate to 0.013475 BTC/day from a single miner.  The group buys are going for about 2.15BTC.

The breakeven point, assuming no difficulty change and no electricity cost is 159.5 days

Factor in difficulty increases and forget it.  In 3 months the difficulty will be so high these will be generating next to nothing.

I don't think that could ever finally make some profit. 

Buying these on ebay is a bad investment. I expect the chance of getting a ROI would be slight, I'd guess at maybe 10%.

Buying them new, for 2.1 - 2.8 is arguably a good investment. I personally don't think they are, but I think the majority believes they are.

The best ROI with these items is in fact purchasing them from the source, and then reselling them, on ebay or your own store.

The product is sexy and has a lot of merits (power,heat, ease of use) but I believe at this juncture, when we sort of between 1st and 2nd gen asics, and difficulty rates with big-papa asics coming online, that there are better investments to be made.

One exception would be perhaps for folks that have live in areas with much higher energy costs than most.

Don't forget that there are EBay fees, Paypal fees, shipping costs, fees to turn that fiat back into BTC.  I've had zero feedback bidders not pay, resulting in more listings, and the fees just keep starting over.   Overall it's a really inefficient way to move these units.

To me, it is clearly way too expensive!

If you look at the price of 2 avalon chips, it will cost you around 0.18 BTC for the chips...

Even if they don't have a good PCB assembly, they win at least 1.2-1.5 BTC per Unit.

But in the end, everybody would do the same: the demand is skyrocket high and they are the only ones shipping.

What the ASIC is going to do is enrich people that had already a lot of bitcoins and were the first to order and receive ASIC. Manufacturers and resellers will also get rich.

All the other ones will pay very expensive rigs and will barely have an ROI.

To me the best thing to do is buy bitcoins through a trading platform and keep them for two months or so.

End of Aug, people will start to realize that unlessnthey spend a lot of money buying expensive ASIC material, they won't profit anymore and start selling their bitcoins. And then... I'm quite sure the BTC exhange rates will drastically go down.

BFL is still vaporware for now, these may not be worth the price but at least they exist

meh they don't look worth it for the price.  might as well wait for a BFL order then buy two of these....

can these be used?

I'm working on blog post, for something about the same cost @200MH/s, but my shipment of boards is stuck..........

They arent worth it...Might as well get a real asic.

Because I don't have any Bitcoins yet. :p I'm not in a position to buy Bitcoins at the moment. I've only got about 0.01 BTC.

Why are you not in a group buy for this, then?

I really don't think more than US$1.00 / Mh/s makes any sense. Basically, you're deciding whether the power savings significantly reduces your cost vs a GPU. Also you are weighing how long it may take you to break even.

This being said, i'm bidding well above this on ebay(and losing) just because i want one in hand. yeah, i'm stoopid.

There is a big difference between LX9 and LX150 (that does about 190-200 MH/s BTW).
And you cannot power LX150 from a USB.

IMO worth about BTC0.2 as you can buy the chips for BTC0.08-0.09 each for assembly on a K16 board

Probably 2 weeks from the moment someone gets the bright idea to spend next couple of weeks writing code for a device to mine (probably) less than 5 MH/s @ $89

IMHO the 5 MH/s and 2 weeks is me being optimistic.

Edit: Corrected typo. the usb thing is $89 not $8

so when is someone going to post the howto make these miners..

http://www.em.avnet.com/en-us/design/drc/Pages/Xilinx-Spartan-6-FPGA-LX9-MicroBoard.aspx

looks to be the same..

They are probably worth 1 to 1.5 coins. Overpriced in my opinion but plenty of people are buying them so they are worth what they are selling for. I personally don't think they will make back their investment though. Although they might if Avalon and BFL continue to falter.

i just dont think they are worth it.

Most recent one on eBay sold for $485 (3.7 btc) but obviously demand will drop as more are offered.

Right now they are worth about BTC2.  Use a local group buy to save on shipping, customs and duties.

I believe the OP was looking for community insight as to it's value, not looking for the best/cheapest price out there.  If I'm wrong I retract my last post and this one.

That price isn't surprising, esp. considering some groupbuy units were almost this high in BTC.  Before a price was even announced I said 30$ - 50$ USD (which based on the assessment for both the public and the company I felt was fair).  Hmm okay either I thought that or said it somewhere else, because I looked through my past posts and this is the only time I mentioned money before the unit was priced...


I'm still hoping that ASICMiner will release more devices, and it would be even better if at least some were in the sub-1000$ USD range (although I'm sure they will sell out anything they produce).  But yeah, it's a little easier (in my mind at least) to say "oh okay, I'm going to pay just 3BTC" compared to "holy crap, there goes $400 from my bank account!", esp. considering a lot of BTC was gained before the price spike (I know 1BTC = ~$270 at one point, however I believe that was a bubble that needed to pop and am more comfortable with the level it's at now... sorry to anyone who got burned though).

There are several group-buys at this forum selling for less than 2.6 BTC...

See for instance https://bitcointalk.org/index.php?topic=195037.new#new or https://bitcointalksearch.org/topic/tentative-asicminer-block-erupter-usb-group-buy-uk-218300... Also, you could use search "USB ASIC"...

I am curious what are people's opinions on the value of one of these USB miners, I know they currently sell on eBay with starting bids of $400.  My opinion is that is due to the scarcity effect.  Once there are more supplies available what does the community think they are worth?