Cryptocoin Mining?

[v8envy]

ArtForz already has an ASIC mining solution. Rumors have him at somewhere around 25% of the bitcoin network.

Another outfit is working on a freestanding mining solution. No words on pricing or capacity, but it's likely an ASIC of some sort. People crunching the numbers on chip production spinup costs peg it at about 5.6 gigahash for $1000-1500. I'm guessing it'll use some low cost SoC and be a miner-worker-in-a-box -- just plug it in, put it on the internet, point it to a pool and you're done.

Once FPGA and ASIC miners become more common I'm convinced the bitcion community and bitcoin usefulness will shrink, not grow -- both have been growing because anyone with an ATI card can get in on the fun. You'll see adoption plummet when in order to play with bitcoins you will need an expensive, single purpose piece of hardware. And bitcoins as a currency are just not viable yet.

TL;DR - GPU mining and bitcoins both should still be around for another 4-8 weeks. After that it's a big unknown.

 

[gorobei]

i've read the asic thread. they are at the "wouldnt it be cool" stage and everything is strictly theoretical and rose colored glasses. a few of them are crunching numbers but no one is looking at long term or counter reactions in the market. at best it will take 2.5 million dollars and 1 year to get it started and finished assuming nothing goes wrong. only one or two of them have even considered what the 28 tera hashes unleashed by the asic standalones will do to difficulty.

I agree that it is bad for the community. as soon as all hashing is in the hands of a few people, bitcoin will die as a populist movement. right now mining is open to everyone with a decent video card. every new miner becomes a bitcoin user and potential speculator with a vested interest in bitcoin as a currency. as soon as only the few people willing to invest 7-10k$ in asic equipment are the only ones mining the future of bitcoin becomes tenuous. if you cant mine easily: current miners will cash out, market volume will drop, confidence in coins will drop and the future of btc will be seriously in doubt.

anyway they are still a year away from ever turning out a product. if they wait to see what SI can do before finalizing designs, that's another 6 months delay. and by the time they do get it out, we will have long since reached the 200,000 block when the 50 coin reward goes down to 25.

 

[v8envy]

I think you're too optimistic on timeframes. ArtForz has already spent the spinup costs to do this, so for him the next step of ramping volume and releasing the chips to the public will not have the 2.5 million cost. I find the estimate of 500k and a higher per-chip (read: $100 vs $28) more believable than the 2 million figure.

The SHA accelerator would be just as viable on a 90nm process as it would on 28. Plenty of capacity on the older processes. So what if it runs at half the clock speed and has half the pipelines per chip if the cost can be reduced by an order of magnitude or two? Complexity of this is lower than the old PhysX accelerator, on the same process. Would you buy a 1.25 gigahash card that fits in a PCI slot and doesn't require multiple PCIe power connectors for $200? I know I'd do that in preference to another GPU.

While the open source ASIC/FPGA miner thread is blue sky the closed source, proprietary solution has been promised in the Sept/Oct timeframe. I find this very likely. Mining with an enormous advantage is only possible for one entity, so long as he stays a relatively small % of the overall network capacity. If difficulty goes up an order of magnitude without an order of magnitude increase in price the entity owning 75% of the network will be selling coins to an ever shrinking pool of others. I can mine 100% of my own block chain to guarantee transactions on my private network and nobody would use it. Which means selling that hardware (and possibly using a big chunk of it to stimulate interest) is the only way to keep the pyramid standing.

 

[gorobei]

everyone in the thread seems to think artforz will keep his chips to himself rather than release.
it is kind of pointless to debate since joel seems to indicate there is a way to block asic mining by altering the bitcoin client/model with a memory requirement. a 128 mb memory minimum to mine would make all the current asic/fpga useless. so if the bitcoin controlling board/group/whatever decide that asic oligopoly is bad, they can revise the model to videocard only with the mem requirement. it all depends on how the board sees it.

 

[YoungGun21]

everyone in the thread seems to think artforz will keep his chips to himself rather than release.
it is kind of pointless to debate since joel seems to indicate there is a way to block asic mining by altering the bitcoin client/model with a memory requirement. a 128 mb memory minimum to mine would make all the current asic/fpga useless. so if the bitcoin controlling board/group/whatever decide that asic oligopoly is bad, they can revise the model to videocard only with the mem requirement. it all depends on how the board sees it.

Why and how would they do that though? They can't just go and change it like that. If he solo-mined, he would just have to have a version of the miner that didn't have that requirement.

 

[gorobei]

Why and how would they do that though? They can't just go and change it like that. If he solo-mined, he would just have to have a version of the miner that didn't have that requirement.

i'm assuming he's talking about changing the hashing algorithm and not some sort of drm tack on. so they change the math used for hashing to require keeping a certain amount of data active on buffer in order to process the block data and the data is large enough (128 mb or larger) to make adding that many memory chips to a asic board prohibitively expensive to manufacture. since video cards already come with ram they wouldnt be affected but as a small run manufacture the asic builder wouldnt be able to get any volume discount on the ram or have room on the pcb to feed 6 cores.

 

[YoungGun21]

But he can't change the bitcoin client that they already have. So in effect, it just prevents new people from taking advantage of it. The big "flaw" (it was purposefully designed this way) is that no person can take control of bitcoin and change all of the rules. The hashing algorithm can be changed from now on for all the new bitcoin client releases, but that won't affect the old ones. So all a person wishing to use an asic board would need to do is install an older bitcoin client, which is readily available online.

 

[Chiropteran]

I've decided to cash out again, traded 44.37 bitcoins for a samsung galaxy 10.1 32gb tablet. been wanting a tablet for awhile, and since bitcoins seem stable for the last few weeks I feel safe enough selling some at this price.

Still mining with my existing rigs, but I don't think my home's electrical circuit can handle anything more so it wouldn't be useful to buy any more mining hardware.

Of course, since I did cash out, expect BTC to peak at $40 in the next week

And of course, see this post 5 years from now, when 1 BTC is worth $25,000, and laugh about my overpriced tablet.

 

[gorobei]

But he can't change the bitcoin client that they already have. So in effect, it just prevents new people from taking advantage of it. The big "flaw" (it was purposefully designed this way) is that no person can take control of bitcoin and change all of the rules. The hashing algorithm can be changed from now on for all the new bitcoin client releases, but that won't affect the old ones. So all a person wishing to use an asic board would need to do is install an older bitcoin client, which is readily available online.

you understand that the bitcoin.exe/wallet-client is separate from bit-miner.exe, phat, phoenix, or poclbm. the mining function was removed from the wallet client a while ago. the hashing process is done by a different program.

if the bitcoin overseeing group changes it so that the block chain requires more than the prior hash(say 5 past hashes or whatever raises the memory required) and a nonce for verification then having an older wallet client isnt going to override that. blocks chain branches that dont comply would be orphaned an reprocessed into the main(128+ memory min required hashing) block chain.

joel isnt saying they will, just that they can.

 

[tommo123]

I've decided to cash out again, traded 44.37 bitcoins for a samsung galaxy 10.1 32gb tablet. been wanting a tablet for awhile, and since bitcoins seem stable for the last few weeks I feel safe enough selling some at this price.

Still mining with my existing rigs, but I don't think my home's electrical circuit can handle anything more so it wouldn't be useful to buy any more mining hardware.

Of course, since I did cash out, expect BTC to peak at $40 in the next week

And of course, see this post 5 years from now, when 1 BTC is worth $25,000, and laugh about my overpriced tablet.

how did you cash out?

 

[DougoMan]

I'm having the strangest problem, whereas the -v -w128 -f0 flags are actually slowing my mining down.

I was doing a steady 420 MHash/s with the -v -w128 -f0 flags on my 6970, but all off a sudden it plummeted to around 300 and stayed there.

Yet strangely if I remove the flags, I actually get a noticeable improvement and manage 370 MHash/s.

This happened right after switching power supplies, so I do not know if there is a connection. I don't see how there could be.

Has anyone encountered something like this?

edit: nevermind, I guess it was throttling itself due to overheating. Weird since it was set to manage the fans itself and was only running at around 40%. I wonder why it did not ramp up to the fans to cool itself.

 

[Chiropteran]

how did you cash out?

I used http://spendbitcoins.com/ for an amazon.com giftcard.

 

[YoungGun21]

I'm having the strangest problem, whereas the -v -w128 -f0 flags are actually slowing my mining down.

I was doing a steady 420 MHash/s with the -v -w128 -f0 flags on my 6970, but all off a sudden it plummeted to around 300 and stayed there.

Yet strangely if I remove the flags, I actually get a noticeable improvement and manage 370 MHash/s.

This happened right after switching power supplies, so I do not know if there is a connection. I don't see how there could be.

Has anyone encountered something like this?

edit: nevermind, I guess it was throttling itself due to overheating. Weird since it was set to manage the fans itself and was only running at around 40%. I wonder why it did not ramp up to the fans to cool itself.

The fans for ours never went above 40% on auto, so we just made them to stay at 80% in order to make sure we had no issues.

 

[wbynum]

With the new difficulty increase I am down to ~1.7 BTC per day for my 3.6 GH/s. At my 10.5 cent kw/h electricity rate and a ~$13 BTC price that is only ~$10 in profit per day. Going to have to seriously think about shutting down my farm soon.

 

[Borat]

deleted

 

[hdfxst]

With the new difficulty increase I am down to ~1.7 BTC per day for my 3.6 GH/s. At my 10.5 cent kw/h electricity rate and a ~$13 BTC price that is only ~$10 in profit per day. Going to have to seriously think about shutting down my farm soon.

and with coins staying at 13-14 dollars i decided to stop at 25 coins,right now i'm at 22.7.I'll cash in my coins and sell my 5830 and 5850 and maybe get a 6950

 

[dajeepster]

I'm having the strangest problem, whereas the -v -w128 -f0 flags are actually slowing my mining down.

I was doing a steady 420 MHash/s with the -v -w128 -f0 flags on my 6970, but all off a sudden it plummeted to around 300 and stayed there.

Yet strangely if I remove the flags, I actually get a noticeable improvement and manage 370 MHash/s.

This happened right after switching power supplies, so I do not know if there is a connection. I don't see how there could be.

Has anyone encountered something like this?

edit: nevermind, I guess it was throttling itself due to overheating. Weird since it was set to manage the fans itself and was only running at around 40%. I wonder why it did not ramp up to the fans to cool itself.

I've got a weirder problem than yours... my 6970 dips down to 7Mhash/s when ever i play mobster2 on facebook... wtf is up with that?!!?.. and i keep my fans at 100%

 

[Muyoso]

and with coins staying at 13-14 dollars i decided to stop at 25 coins,right now i'm at 22.7.I'll cash in my coins and sell my 5830 and 5850 and maybe get a 6950

Did you ever crossfire those two? I have the same setup and was planning on trying crossfire for games when I am done mining.

 

[hdfxst]

Did you ever crossfire those two? I have the same setup and was planning on trying crossfire for games when I am done mining.

I never ran crossfire before but i did try it on these two cards and had nothing but lockups.Now that was with the old trixx where both cards were being clocked to the same speeds so i don't know if that was it or the 50 dollar xfire motherboard that i'm using,I think it was a combination of my inexperience and this cheap mobo(the 5850 would not work in the x4 slot).I figure i can sell these 2 while they're still worth something and end up with 6950 or 6970 for basically nothing

 

[YoungGun21]

Selling my 3 Sapphire 5830s and two 750w PSUs, if anybody is still looking to put money in to this!

Moving in to my own apartment, don't want to deal with a huge electricity bill. Cards run great.

 

[gorobei]

and with coins staying at 13-14 dollars i decided to stop at 25 coins,right now i'm at 22.7.I'll cash in my coins and sell my 5830 and 5850 and maybe get a 6950

i'm in roughly the same boat, i'll probably mine for another month at a target of 30 coins and stop. If SI is a big enough jump in hashes , i might stay in a little longer.

 

[dalamchops]

i'm in roughly the same boat, i'll probably mine for another month at a target of 30 coins and stop. If SI is a big enough jump in hashes , i might stay in a little longer.

it's worth it to mine through the winter for me. Can be used as heat.

 

[hdfxst]

I figured it up today and i would have to run 4 5830's to equal what my single 5850 was getting when i started this

 

[3DVagabond]

i'm in roughly the same boat, i'll probably mine for another month at a target of 30 coins and stop. If SI is a big enough jump in hashes , i might stay in a little longer.

Do you think SI will be a jump in hashes? I was thinking it might be the opposite with fewer, but deeper, threads. That's if I understand the GCN arch correctly. I might not.

 

[sm625]

This is kind of a retarded question, but has anyone had their guiminer window simply refuse to display itself? It shows on my taskbar but it just wont display (maximize or restore) the actual window no matter what I do. It used to work just fine. The older version still works ok.