Is BFL causing the difficulty increase?

[norulezapply]

I've noticed the price seems to be more or less level at the moment, but difficulty is steadily increasing. And this seems to have been the trend since everyone found out the BFL Singles are actually real..

Is this difficulty increase due to so many people ordering BFL Singles?
After reading through https://bitcointalk.org/index.php?topic=67887.0, it made me realise just how popular these things are, with some people ordering 10+, which is gonna make difficulty shoot higher surely.

Are GPUs finally becoming obselete in the next couple of months as people get their BFL Singles delivered?

What's your opinions on this?

[Littleshop]

I doubt it.  Solominer 88.6.216.9 is doing more to cause difficulty increase then any other single source.  They may be using BFL (I doubt it), but if they keep increasing at the rate they are going I would say they are using ASIC. 

[PulsedMedia]

I've noticed the price seems to be more or less level at the moment, but difficulty is steadily increasing. And this seems to have been the trend since everyone found out the BFL Singles are actually real..

Is this difficulty increase due to so many people ordering BFL Singles?
After reading through https://bitcointalk.org/index.php?topic=67887.0, it made me realise just how popular these things are, with some people ordering 10+, which is gonna make difficulty shoot higher surely.

Are GPUs finally becoming obselete in the next couple of months as people get their BFL Singles delivered?

What's your opinions on this?

GPUs still buy more Mhash / $/€/BTC, while having lower efficiency.
I see GPUs being ran side by side with FPGAs, those who have low electricity rates can operate GPUs quite profitably, while those with high priced electricity will opt for FPGAs.

ASICs however is going to be an killer application, a game changer most likely.

[Dhomochevsky]

The difficulty will not increase for the next step (tomorrow morning or at noon the next difficulty step will begin). At most, you'll see a 1-2% change which can go either way - right now it's estimated it will go down 1%, but that can change. Difficulty for the next step will be pretty much the same as it is now.

ASIC's are far too big of an investment for many people to make, at least at this time, when FPGA mining is just starting to take off. Make no mistake, ASIC's are THE end-game for Bitcoin mining. Eventually, a few miners that manage to invest enough money in ASIC's will grab most of the hashing power of the network, but right now we need to pass through the FPGA adoption era.

GPU's WILL become obsolete, but I don't think it will be in the next couple of months. It will take a little bit longer, but by the end of 2012, especially with the reward halving, I estimate that very few people will find that mining with GPU's is profitable.

[edit] BFL is most definitely not causing any difficulty disruptions. They shipped far too few units so far to make even a dent in the hashing power.

[kakobrekla]

If you ask me, .... x6500, icarus, ztex... each of them alone delivered more MH/s to customers than BFL.

[Matthew N. Wright]

If you ask me, .... x6500, icarus, ztex... each of them alone delivered more MH/s to customers than BFL.

Probably. The thing is, just like Apple products, people will care about simplicity, design, etc much more than a few Mh/s.

Butterfly Labs will be huge, and I am looking forward to working with them on a collaborative effort later with our BitDex.

[norulezapply]

Thanks for the input.

Correct me if I'm wrong but doesn't the BFL Single give the best MHash/$ currently out of all the available FPGAs? In my opinion that'd be the biggest driving factor for purchasing them over any other current FPGA, even if they are quite inefficient in comparison, they are still much more efficient than GPUs.

Also I know not many have shipped yet but it looks like there's a large order book which they will eventually get through, so I guess they'll have the main hash rate share of all FPGAs once the orders are delivered.

Also Matthew, what's the BitDex? The first thing that came to my mind when I read that was PokeDex...

[this time]

A bfl single is about 2x as efficient as a gpu. An icarus/ztex is 2x as efficient as a bfl single.

[Pipesnake]

I'm sure the less-than-20 BFLs that are actually out in the wild are making a HUGE difference.   

[RoloTonyBrownTown]

A bfl single is about 2x as efficient as a gpu. An icarus/ztex is 2x as efficient as a bfl single.

Mh/w yes.   Mh/$, no.

[Kluge]

If you ask me, .... x6500, icarus, ztex... each of them alone delivered more MH/s to customers than BFL.

Probably. The thing is, just like Apple products, people will care about simplicity, design, etc much more than a few Mh/s.

Butterfly Labs will be huge, and I am looking forward to working with them on a collaborative effort later with our BitDex.

[cunicula]

I doubt it.  Solominer 88.6.216.9 is doing more to cause difficulty increase then any other single source.  They may be using BFL (I doubt it), but if they keep increasing at the rate they are going I would say they are using ASIC. 

Don't worry about Solominer causing a sustained difficulty increase. If he keeps growing, the GPU miners will begin to drop out, preventing further rises in difficulty.

[cunicula]

the GPU miners will begin to drop out

Maybe, but I wonder how many GPU miners are in the same position as me. I've paid off my GPUs ages ago. Electricity is cheap. It's silly how low the price has to go or how high the difficulty has to rise before I'm mining at a loss.

When my GPUs are no longer profitable, I'll sell them and replace them with as many FPGAs as the GPUs can buy (it helps I've been swapping out 5xxx for 7xxx in the process). Now, since those FPGAs will be already paid for, how long can I mine with them before it becomes unprofitable? Sure it will be less hashing power than my GPUs but FPGAs will be so cheap to run that the difficulty can hit ridiculous levels and I'll still be making money.

Mining has been profitable for so long now, I don't expect any massive decrease in the global hash rate even with an increasing difficulty combined with a halved block reward.

You could be right. I really don't know. Where I live, electricity alone would put me just at the break-even point now. Given additional costs due to cooling and familial harmony, price would have to go up to about $9 for me to turn on the GPUs again.

[Matthew N. Wright]

Also Matthew, what's the BitDex? The first thing that came to my mind when I read that was PokeDex...

Hehe. Exactly. Think Pokedex, but for Bitcoin (and a lot smaller, sleeker, and it doesn't talk to you everytime something happens).

BitDex is just a codename. We aren't releasing the real name until the product is ready for production for obvious reasons.

[norulezapply]

I'm sure the less-than-20 BFLs that are actually out in the wild are making a HUGE difference.   

some people ordering 10+, which is going to make difficulty shoot higher

Read the original post please...

Exactly. Think Pokedex, but for Bitcoin

This sounds interesting... Some kind of business directory/general knowledgebase for Bitcoin I assume. But most of that is already available on the Wiki anyway. Hmm..

[P4man]

ASIC's are far too big of an investment for many people to make, at least at this time, when FPGA mining is just starting to take off.

Search for LargeCoin on this forum. May or may not be legit, but it appears someone has bit the bullet already.
Asics (and to a lesser extend, s-asics) have been the prime reason I have not committed to FPGAs. At least GPUs I can resell to gamers, but its gonna be tough reselling your 80W BFL single's if asic based products really hit the market.

[rjk]

ASIC's are far too big of an investment for many people to make, at least at this time, when FPGA mining is just starting to take off.

Search for LargeCoin on this forum. May or may not be legit, but it appears someone has bit the bullet already.
Asics (and to a lesser extend, s-asics) have been the prime reason I have not committed to FPGAs. At least GPUs I can resell to gamers, but its gonna be tough reselling your 80W BFL single's if asic based products really hit the market.

Largecoin was confirmed by ttul to be s-asic, if I remember correctly.

[P4man]

Largecoin was confirmed by ttul to be s-asic, if I remember correctly.

I believe the current claim is that its a custom asic, but of course, whats in a name, a structured asic like fast path I guess is also "custom" and .. application-specific, and its an integrated circuit.. so, it wouldnt be a real lie.

s-asic always seemed the obvious way to go to me, I was even convinced thats what BFL did. Compared to fpga, it seriously lowers per unit cost with reasonable volumes, it has good potential to lower power consumption considerably and the NRE is not nearly as extreme as with a full custom design.

[rjk]

Largecoin was confirmed by ttul to be s-asic, if I remember correctly.

I believe the current claim is that its a custom asic, but of course, whats in a name, a structured asic like fast path I guess is also "custom" and .. application-specific, and its an integrated circuit.. so, it wouldnt be a real lie.

s-asic always seemed the obvious way to go to me, I was even convinced thats what BFL did. Compared to fpga, it seriously lowers per unit cost with reasonable volumes, it has good potential to lower power consumption considerably and the NRE is not nearly as extreme as with a full custom design.

As I read the ttul post it pretty much says that the current offering is based on sASIC.

One day someone will build a standard cell based ASIC for Bitcoin mining, and that will cost $5-10M NRE and it will be a monster.