Pages:
Author

Topic: [Poll]Will you be switching to ASIC? (Read 7207 times)

newbie
Activity: 58
Merit: 0
August 28, 2012, 07:01:02 PM
#72
I think there's a lot of angle's here but this is the evolution of bitcoins.
ACIS were not made specifically for bit coins
they existed before., bitcoin is not a get rich quick scheme
so even if the profit so low for GPU when ASIC drops
it depends on the person if they want to upgrade.

Also one big thing butterfly labs is an american company
and this product at 1333.00 and 30g's we aren't talking cheap
they are helt over a standard and at 6000+ order i don't
think they plan on disappointing they have come through
with there mini rig's in the past., so the logical thing if
your going to keep mining is to go in with ASIC 
newbie
Activity: 8
Merit: 0
August 28, 2012, 10:24:12 AM
#71
i really wonder how many gpu miners will be switching to LTC when asic comes out.  a couple posts mentioned it, but LTC could really take off if all the current (minus fpga) hashing power went over to LTC.  perhaps another poll is in order.
sr. member
Activity: 364
Merit: 250
August 16, 2012, 04:17:56 PM
#70
Well, I could say you have to have a banking license, 'just to participate' in banking business. You need a license for frequency 'just to participate' in communication business. And there are no small licenses for starters, i.e. not affordable for an average person. Bitcoin personal banking system lets you 'mine' at very low cost. Right now you're complaining that in order to dig for gold you have to buy a shovel, since all gold on the surface is gone and you can't just take it with your hands, so to speak. However, unlike shovels for gold digging, ASICs can be eventually integrated into other, more casual items, like cars, public bitcoincard processing hubs, even laptops and cellphones. That's if Bitcoin becomes widely accepted....

I think trying to compare mining bitcoin to two federally regulated industries is a bit of a straw-man argument. With that said I'm not even going to bother talking about those points.  As per your shovel and mining gold analogy:  I view digging with your hands akin to CPU mining and the shovel akin to GPU mining.  You can mine for gold with a shovel, sure. You can also pot plants; dig a grave, ditch, hole; and clean up dog shit.  A shovel is a pretty versatile tool that can be used for many other things than digging for gold.  In fact, a lot of people already own shovels so re-purposing the shovel to dig for gold means anyone who owns a shovel can potentially dig for gold.  I suppose one could say that the FPGA is like a back-ho.  It's much more efficient, however it costs more to purchase and at the end of the day, in most places you need to be a ticketed operator to use one.  I view the ASIC kind of like a specialized industry specific machine:



 It's going to be insanely efficient at serving it's purpose (ignoring cost), and it's not something you have just kicking around the backyard.  It's also got no real use outside of it's intended industry.


Good thinking, love this analogy.  I am still very skeptical of ASIC though.. we will see. I don't know much about mining, though. Just what I learn inexorably.
member
Activity: 62
Merit: 10
July 20, 2012, 03:02:38 PM
#69
I think the BFL ASIC numbers could be close to accurate...If we look at academic research we can see that 5x speedup over FPGA is feasible on ASIC.  In fact, considering the type of algorithm that mining requires there are several papers documenting similar algorithms achieving a 4-5x speedup on ASIC over FGPA.

Now crunch a little numbers...

BFL puts two 2 FGPA  in BF single, so that's ~415 Mhash/s per FPGA.  Assume they can achieve the maximum speedup, 5x, over their FGPA = 2.075 Ghash/s per chip.  

One chip won't reach the 3.5 GHash/s reported for Jalapeno, so it means Jalapeno must have 2 ASICS (and also that they are probably not achieved the 5x speedup), at $149 per Jalapeno thats ~$75 per chip.

Single SC is at 40Gh/s that's ~19 (lets just say 20 because probably not exactly 5x speedup) chips that's $1299/20 = $61.00 per chip

Mini rig at 1Th/s is ~500 (!) chips.  At $30000 that's $60.00 per chip


Now the question is if they can raise enough funds to cover the total production cost (~12 mill for $26.00 per chip using the estimation I found) and deliver by October.

This is of course just back of the envelope calculations, but it does show that their numbers are feasible..

legendary
Activity: 1204
Merit: 1000
฿itcoin: Currency of Resistance!
July 18, 2012, 03:18:56 PM
#68
I think I'll pre-order a 2~4 Jalapenos soon...
But don't know yet...
I'm still buying some 7970...
member
Activity: 75
Merit: 10
July 18, 2012, 03:16:34 PM
#67
I own GPUs, and I'm switching to BFL ASIC, i choose this one Grin
hero member
Activity: 628
Merit: 504
July 18, 2012, 08:35:15 AM
#66
There are only 7200 coins to be mined every day no matter what kind of mining machine people setup, this is a big difference compared to gold mining

Newer hardware and heavy investment just redistribute a portion of the daily coin to people who have higher hashing power, and unless the BTC value increase exponentially, their return on investment will drop quickly as they invest more, I think this is the most interesting part in BTC


You're right, ultimately the demand for investment is dictated by exchange rate , and thus the difficulty too.
legendary
Activity: 1988
Merit: 1012
Beyond Imagination
July 17, 2012, 03:11:08 AM
#65
There are only 7200 coins to be mined every day no matter what kind of mining machine people setup, this is a big difference compared to gold mining

Newer hardware and heavy investment just redistribute a portion of the daily coin to people who have higher hashing power, and unless the BTC value increase exponentially, their return on investment will drop quickly as they invest more, I think this is the most interesting part in BTC

hero member
Activity: 810
Merit: 1000
July 16, 2012, 01:06:01 AM
#64
so this survey is a bit useless


An interesting question that I hope this poll will answer is what existing non-BFL FPGA miners will do.

Will they sell their device on the secondary market to recoup costs or continue mining because the incremental cost to do so is next-to-nothing?



If ASICs hit the market, I will be reprogramming my FPGAs to undertake other business work such as LTC mining, cracking lost passwords as a commercial spin off or some other number crunching venture.

This sort of thing is not for the average Joe but then again I am an electrical / electronics engineer whom should be able to pick it up over a few weeks.
hero member
Activity: 628
Merit: 504
July 16, 2012, 12:10:37 AM
#63
Well, I could say you have to have a banking license, 'just to participate' in banking business. You need a license for frequency 'just to participate' in communication business. And there are no small licenses for starters, i.e. not affordable for an average person. Bitcoin personal banking system lets you 'mine' at very low cost. Right now you're complaining that in order to dig for gold you have to buy a shovel, since all gold on the surface is gone and you can't just take it with your hands, so to speak. However, unlike shovels for gold digging, ASICs can be eventually integrated into other, more casual items, like cars, public bitcoincard processing hubs, even laptops and cellphones. That's if Bitcoin becomes widely accepted....

I think trying to compare mining bitcoin to two federally regulated industries is a bit of a straw-man argument. With that said I'm not even going to bother talking about those points.  As per your shovel and mining gold analogy:  I view digging with your hands akin to CPU mining and the shovel akin to GPU mining.  You can mine for gold with a shovel, sure. You can also pot plants; dig a grave, ditch, hole; and clean up dog shit.  A shovel is a pretty versatile tool that can be used for many other things than digging for gold.  In fact, a lot of people already own shovels so re-purposing the shovel to dig for gold means anyone who owns a shovel can potentially dig for gold.  I suppose one could say that the FPGA is like a back-ho.  It's much more efficient, however it costs more to purchase and at the end of the day, in most places you need to be a ticketed operator to use one.  I view the ASIC kind of like a specialized industry specific machine:



 It's going to be insanely efficient at serving it's purpose (ignoring cost), and it's not something you have just kicking around the backyard.  It's also got no real use outside of it's intended industry.

All of those points aside, I think there's a massive disconnect between the current ASIC state and having manufacturers designing BTC mining ASIC into everyday appliances.  Do you honestly think you're going to see GM shipping trucks that mine bitcoins while you drive at any point, ever?  

Yes, you've provided a better analogy. But imagine being able to have that huge machinery (or a small one) which does it's work without you having to drive it or do anything? Just plug-n-...work Smiley. That's what I'm trying to get through here. I have no idea if we ever see cars with ASICs on board, just like I had no idea if we will ever see a car with tv, computer, hybrid power and etc... Time will show, sorry for using cliches.
legendary
Activity: 1512
Merit: 1000
July 15, 2012, 06:17:26 PM
#62
Well, I could say you have to have a banking license, 'just to participate' in banking business. You need a license for frequency 'just to participate' in communication business. And there are no small licenses for starters, i.e. not affordable for an average person. Bitcoin personal banking system lets you 'mine' at very low cost. Right now you're complaining that in order to dig for gold you have to buy a shovel, since all gold on the surface is gone and you can't just take it with your hands, so to speak. However, unlike shovels for gold digging, ASICs can be eventually integrated into other, more casual items, like cars, public bitcoincard processing hubs, even laptops and cellphones. That's if Bitcoin becomes widely accepted....

I think trying to compare mining bitcoin to two federally regulated industries is a bit of a straw-man argument. With that said I'm not even going to bother talking about those points.  As per your shovel and mining gold analogy:  I view digging with your hands akin to CPU mining and the shovel akin to GPU mining.  You can mine for gold with a shovel, sure. You can also pot plants; dig a grave, ditch, hole; and clean up dog shit.  A shovel is a pretty versatile tool that can be used for many other things than digging for gold.  In fact, a lot of people already own shovels so re-purposing the shovel to dig for gold means anyone who owns a shovel can potentially dig for gold.  I suppose one could say that the FPGA is like a back-ho.  It's much more efficient, however it costs more to purchase and at the end of the day, in most places you need to be a ticketed operator to use one.  I view the ASIC kind of like a specialized industry specific machine:



 It's going to be insanely efficient at serving it's purpose (ignoring cost), and it's not something you have just kicking around the backyard.  It's also got no real use outside of it's intended industry.

All of those points aside, I think there's a massive disconnect between the current ASIC state and having manufacturers designing BTC mining ASIC into everyday appliances.  Do you honestly think you're going to see GM shipping trucks that mine bitcoins while you drive at any point, ever?  
member
Activity: 118
Merit: 10
July 15, 2012, 01:59:19 AM
#61
ASICs can be eventually integrated into other, more casual items, like cars, public bitcoincard processing hubs, even laptops and cellphones.

This won't happen because those small mining units will be inefficient (compared to a dedicated mining rig with minimal variable costs) and will likely run at a loss.  You've also got to have storage space for the block chain, a fat enough pipe to handle the bajillion transactions per block in the Bitcoin of the future, cooling and all kinds of stuff that makes mining in anything else than a dedicated setup with low costs unprofitable.
hero member
Activity: 628
Merit: 504
July 14, 2012, 03:24:14 PM
#60

Some people just don't get it. The fear of loosing profit blinds I guess...or maybe just trolling others.

Were you referencing my statement, or someone elses? 
Well, if you're agreeing that ASICs will kill the 'casual' miner, then yes I refer to your statement. What BFL does, is giving miners ability to purchase high efficiency products with high range in price. They effectively cover prices from $150 to $30k. They sell to anyone. Now tell me, why is the casual miner will be killed? I say, the casual miner's thickheadness conservatism could kill him, nothing more, nothing less...

I suppose I view it like this.  The casual miner could have mined coins back in the day just by owning a computer.  Then, you could mine if you had a decent enough video card.  As soon as that was figured out, it essentially killed the ability for anyone who owns a basic computer to participate unless they had a decent video card.  If you had a computer with a decent video card, just turn it on and away you go.  When you're done, go back to your gaming or what have you.  FPGA's came out but didn't really mess with the ability for those with just a decent video card to participate.  But, now with the ASIC  what's been done is essentially made it so that anyone who wants to participate has to spend money they normally wouldn't have on an application specific device that has absolutely 0 purpose outside of mining... just to participate.

 


Well, I could say you have to have a banking license, 'just to participate' in banking business. You need a license for frequency 'just to participate' in communication business. And there are no small licenses for starters, i.e. not affordable for an average person. Bitcoin personal banking system lets you 'mine' at very low cost. Right now you're complaining that in order to dig for gold you have to buy a shovel, since all gold on the surface is gone and you can't just take it with your hands, so to speak. However, unlike shovels for gold digging, ASICs can be eventually integrated into other, more casual items, like cars, public bitcoincard processing hubs, even laptops and cellphones. That's if Bitcoin becomes widely accepted....
full member
Activity: 123
Merit: 100
July 14, 2012, 01:07:47 AM
#59
Running with GPUs currently. When an ASIC for bitcoin mining physically exists I will most likely purchase one. I was discouraged from pre-ordering since BFL does not currently offer any sort of payment option that can be refunded if they breach their promises via release date or performance. Bank wire or BTC are the only payment options, neither of which can be recovered without co-operation of the recipient.

Thralen
rjk
sr. member
Activity: 448
Merit: 250
1ngldh
July 13, 2012, 10:47:10 PM
#58
Isn't that how all the old boards are configured and tuned, using different firmware? You are saying if someone buys a single they could have a $1300 brick on their hands?

What it means is that all the tuning happens at the factory, with the exception of overclocking, if that is an option.
legendary
Activity: 1274
Merit: 1000
Personal text my ass....
July 13, 2012, 04:28:43 PM
#57
Isn't that how all the old boards are configured and tuned, using different firmware? You are saying if someone buys a single they could have a $1300 brick on their hands?
legendary
Activity: 1274
Merit: 1000
Personal text my ass....
July 13, 2012, 04:05:32 PM
#56
wheres are the options "i dont have gpu's or fpga's but i plan on buying BFL-asic or non-BFL-asic"?

This is me also. 
sr. member
Activity: 476
Merit: 250
Keep it Simple. Every Bit Matters.
July 13, 2012, 09:50:15 AM
#55
I'm very much I'll believe it when I see it kinda person. BFL have not proven their latest product, so I have no interest in being pulled in by their marketing out of fear of losing out, on some kind of theoretical edge if and when they do exist. I didn't miss out on any edge when the FPGA's came out, it won't be any different with the ASIC's.
I will however comment on it and watch by the side lines however.

So I will bide my time and wait until someone can actually prove an ASIC is working effectively, efficiently and wants to sell them at a half decent price. Since I just don't think they can pull off all they stated and because of that will wait until their is an ASIC that I do believe will perform like they say.

Creating ASIC is hardly a simple or easy thing, my dad (Semi-retired) did work in the semi-conductor industry and has done his fair share of projects with ASIC and FPGA's, so he let me know about the process of the different ASIC and FPGA many years ago and refreshed my memory just recently when I got into bitcoin. FPGA's are so much easier to work with for a reason, they can be reprogrammed, which is a huge advantage in newer markets. Bitcoin is a new thing, not like their is a bunch of ASIC sitting around that happen to do exactly what we need it to do.

Bitcoins as a new market, could change, security algorithm could be change, entire method could be altered, you just don't know what is around the corner. Also just look at how often with the FPGA's the hardware developers have updated the firmware for them and still are. That won't be an option with an ASIC, if there is a mistake, a bug or low performance your stuck with it.

The design process for creating an ASIC is very intense and long, something I'm not sure BFL is ready for and most definitely underestimating, I also don't think the bitcoin itself is mature enough to know nothing will be changing that won't effect an ASIC hardware 1-2 years from now making them have a short lifespan.

They are actually expensive to create, because of how they are built to specifically for your design, they are generally only viable cheaply individually when you create 100,000's + so I still think they are exaggeration of course on all their numbers, since it's just not viable at a price point let alone comparing the performance numbers. They've had a lot of orders, maybe 500-1000, I know the bigger ones would have multiple chips, but it still doesn't reach the bulk order to warrant such a low price for them.

However their is no doubt ASIC will eventually outperform FPGA's, just like the are starting to with GPU's.
I think it's too early to say ASIC's will be the next big thing and also don't think it be any time soon.
FPGA's and ASIC are common place in many other industries, they aren't always dominant over each other.
hero member
Activity: 700
Merit: 507
July 13, 2012, 05:20:12 AM
#54
I might switch once ASICs exist. But since they dont exist i dont bother thinking about it.
legendary
Activity: 1512
Merit: 1000
July 13, 2012, 04:51:51 AM
#53

Some people just don't get it. The fear of loosing profit blinds I guess...or maybe just trolling others.

Were you referencing my statement, or someone elses? 
Well, if you're agreeing that ASICs will kill the 'casual' miner, then yes I refer to your statement. What BFL does, is giving miners ability to purchase high efficiency products with high range in price. They effectively cover prices from $150 to $30k. They sell to anyone. Now tell me, why is the casual miner will be killed? I say, the casual miner's thickheadness conservatism could kill him, nothing more, nothing less...

I suppose I view it like this.  The casual miner could have mined coins back in the day just by owning a computer.  Then, you could mine if you had a decent enough video card.  As soon as that was figured out, it essentially killed the ability for anyone who owns a basic computer to participate unless they had a decent video card.  If you had a computer with a decent video card, just turn it on and away you go.  When you're done, go back to your gaming or what have you.  FPGA's came out but didn't really mess with the ability for those with just a decent video card to participate.  But, now with the ASIC  what's been done is essentially made it so that anyone who wants to participate has to spend money they normally wouldn't have on an application specific device that has absolutely 0 purpose outside of mining... just to participate.

 
Pages:
Jump to: