Bitcoin Forum

So, obviously it's a little late to jump on the ASIC bandwagon. If I were to buy now, the market would be flooded by the time I got one. So I'm looking to the future.

If current Avalon and BFL(if they ever get around to shipping) are the current gen of ASIC, when do you think they'll start accepting preorders for the next-gen(twice as fast, twice as cheap, if ASIC follows all the other electronics on the market)? I'm thinking next year, but I haven't been around that long and don't know these companies as well as the older members here.

I'm honestly not expecting there to be any concrete facts, I'm just asking your opinions. But I welcome concrete facts too. :D

Hard to believe for me... and why should a new gen be created? Its not even near the border where powercost versus mining-revenue comes into a bad connection. So i doubt it will be soon.
And twice as fast? Plus if you go smaller in density it will go up in price too.

Some guys from Switzerland claimed they had a 28nm ASIC and were looking for system integrators here on the forums some time ago. They said more will be announced in May so you might want to keep your eyes peeled. That's the best ASICs can get right now. And I don't think there will be any physical product incorporating 28 nm tech before late summer/early fall.

That said, the most accesible ASICs right now will be the 110 nm ones (Avalon). With chips starting to be delivered in about three months from now, if you want to buy a mining ASIC your best bet will be to purchase from the various DIY projects that will crop up afterwards. Investing in Friedcat's ASICMiner (buying shares) right now would be ill-advised (at least from my point of view), since that market is highly irrational (I think a share is getting close to 2 BTC right now and the mining blades auction they held recently reached ridiculous bids). So, if you want in on the ASIC mining gear, I think the best decision right now would be to play the waiting game and watch the forum closely for new developments. And don't get scammed by the various fake ASIC projects that keep cropping up.

Thank you so much for the answer. I will undoubtedly keep an eye out for news about the 28nm ASIC, hopefully it will be my chance to get into mining while it's still profitable. Besides, maybe I can make some BTC through trading while I wait. ;)

That is how I am making my BTC as well. I think the future may be moving away from mining and simply trading.

In GPU mining era, even there were thousands of users running GPU and FPGAs, you still can get some decent amount of coin if you have a GPU mining rig. We are still in early stage of ASIC mining

You best choice is probably avalon batch 4, or  i like to say mini avalon. They say june delivery. bfl order will be like october or never as some people on here say.

Best bet is the DIY game, otherwise you will have to wait. It's not too late to buy into the ASIC chips if you want to do DIY, Avalon will continue to sell them.

As far as "next-gen" who knows what BFL will do, and Avalon is tricky to deal with, since I'm not willing to pay up that kind of money for a Avalon unit. BFL may just sit on their money a while and snicker at the fact they've had us on our toes for nearly a year.

When next-gen chips do come out, if they don't upgrade their manufacturing tech, they'll at least double the output of their devices for the same price.

In 6 to 7 months the ASIC market will have reached equilibrium and be barely profitable unless the price of bitcoin has risen dramatically.

The only way a next gen ASIC could make money would be if it provided more than a 64x increase in speed for the same power input. I think it is more likely to be 16x and use more power.

Consider, right now, (@ 0.142 kWh)

1 GH/s uses 6.6W and returns 1.6 btc/mo. = $196.99 - 0.67 elec. = $196.31 profit.

Once difficulty has risen 100x (at current trend in 6-7 months),

1 GH/s uses 6.6W and returns 0.02 btc/mo. = $1.97 - 0.67 elec. = $1.30 profit.

Now increase speed 16x (since 28nm * 4 ~ 110nm, very crude estimate),

16 GH/s uses 6.6W and returns $31.52/mo. - 0.67 elec. = $30.84 profit.

That's simply not enough profit to make the investment worthwhile and it would be even less if the power use cannot remain the same for a 16x increase in speed. It approaches diminishing returns unless a next gen can produce 64x speed for same power.

For this reason I think if you don't get on this now then you will not do it later.

how about a Tera hashing machine?
https://bitcointalk.org/index.php?topic=179033.0

Yes, I imagine so, along with the difficulty rising. However these people investing should make most of that money back very quickly at the current difficulty level. And at 6-7 months they would still technically profit by day, but not bank most likely. Much like GPU miners today who were banking before the difficulty level rose.

I'd buy asic now only if I could get it in a week or two, everything else is the risk.

Thought about buying chips and paying someone for assembling but it's still some 4 months time, we have no idea where we will be at that time.

I wish asics never appeared, it was much more fair and long run game for bigger group of people. Asics are to bitcoin mining what Walmarts are to small neighborhood stores - game killers.

I ordered some chips. I'm really hoping they aren't pushed back and arrive in more like 2-3 months.
My 20 chips would earn $1100/mo now. I'm planning on 3x difficulty when they arrive so that should end being $366/mo. At least this long wait should allow getting everything designed and built ready for arrival. I'll be designing/building myself so at least I won't be waiting on another party to assemble and ship a second time, as many will.

With BFL on the verge of shipping difficulty will likely be higher.

Here's some short term analysis of the curve:
https://bitcointalk.org/index.php?topic=178051.0

well, if that comes to be true those who ordered chips are going to have hard time to meet a roi anytime soon.

Most ASIC buyers will absolutely never recover their investment.  And so why would anyone buy one now if they know they difficulty is about to skyrocket?

That seems to be about what I expected. If you're not planning for at least doubling each month over the coming months then you may be in for a surprise. I'd be really pissed if I were a BFL pre-order holder and didn't get my rig for another few months.

I think most of us who ordered chips are already expecting a doubling each month. Maybe I'm wrong and everyone else has their head in the sand. My cost for chips and making boards will be a third what Avalon batch 2,3 paid so I'm pretty happy with the risk on this. If it doesn't recoup costs in 1 month it will in 2 or 3 and that's better than with my GPUs. Everyone said they wouldn't break even in 2011 but I'm still mining now and it's more profitable then ever.

It's funny how people run around saying bitcoin will be over $1000 one day and others run around saying you'll never break even on mining. The two are mutually exclusive.

dOUBLING IS OK BUT THAT PREDICTION OVER THERE ON THAT LINK (damn caps lock) expects some 16 times harder difficulty in a time asics made from chips will start working. If that happens, that investing won't meet roi soon if ever.

correct... ASIC just killed bitcoin... follow the link in my sig to help us crack scrypt!

It's far more profitable and power-efficient to just create a new account and start scamming.  BTW, would anybody be interested in an a couple Avalon batch 2 pre-orders? 100 BTC ea :P

classic!

Not sure how you got that from his table. I see 4x increase by July 19th, almost 3 months from now. Which is close to what I expected when I went ahead.

But there are many variables and any option is a gamble and always has been. I expected to shut down my GPUs last November but delays in ASICs and bitcoin price increases rather surprisingly has made GPU mining more profitable now than even 2 years ago. So, just never say never. As long as people aren't thinking difficulty will drop or doesn't matter they have their own choices to make regarding the future.