Topic: What'll happen to ASICs when they're no longer profitable to run? (Read 3186 times)

Highest quality post in a good while.
Thank you, DnT.

Ente

Maybe. My electricity is expensive. New tech is cool.

Except in reality probably not.  Miners have generally not preferred technology which has a higher capital cost and marginally lower operating cost.  I don't see that changing anytime soon.   A company dumping millions (plural) into 20nm NRE knowing their marginal cost of production is higher than 28nm counterparts would be taking a massive risk.  The only way to put that risk on consumers would be through pre-orders.  I don't really see pre-orders that cost MORE (per MH) than available gear being popular.

If today you could get a miner from one company at $10 and 1 W per GHash with delivery tomorrow OR you could pre-order one from another company for $15 and 0.5W per GHash with delivery in 60 to 75 days would you really pick the latter?

You have missed something rather important there.

A smaller transistor will switch faster and use less energy to switch than a larger transistor. So, a straight reduction of an existing design from 28nm to 20nm ought to give more hashes for fewer watts and therefore it would be a preferred technology even if the transistor prices were higher.

However the smaller stuff can be pumped out at an even lower price point.  Maybe $2 per GH/s.  The raw silicon costs have the potential to be the minority of the overall system. Take HF chip. 19mm x19mm @ 28nm.  If you assume $10K per 300mm wafer it works out to ~$0.15 per GH/s.  Granted you still have testing, cutting, packaging, etc so lets say $0.25 per GH/s. However say HF (or Bitfury or KNC) decides to just sell raw chips in bulk @ $2 per GH/s and system builder builds a rig for another $1 per GH/s. and marks it up a buck for say $4 per GH/s.  That is getting close to the build cost @ 110nm.  Now consider that even if you could buy a rig from either process for $4 per GH/s.  Why would you buy then one which uses almost 12x the power for the same hashrate.

The good news is that ASICMiner and Avalon should have plenty of cash to tape out a smaller process chip, something in the 28 to 55 nm range.

That will probably work to some extent, if you look at the profitability charts, you will see a LOT of scrypt based alts that will get you more than twice the BTC per day that just mining BTC will. The only main offerings on SHA256 alts have been PPC and TRC, not really successful, but ZET is looking ok, it 143.54% BTC on Coinchoose atm.

ZET just listed on Cryptsy a few hours ago, so it's still finding it's market price. If you are going to mine it don't dump insane amounts of ASIC onto it and screw the block chain like TRC, bring it up gradually if you want to preserve the price of the coin.



   

Someone who is heavily staked in low GHash ASICs will try and create a new SHA Alt coin, to attract all of the people who were left behind by the next gen asics. Same thing happened with GPUs and 1st Gen asics. The Scrypt Alt coin craze began to draw in all of those BTC Miners who didn't want to sell off their GPUs at next to nothing to buy promises of ASICs that would come in a year.

It would be closer to 50TH/s I think. If you add up the hash rates of the scrypt coins on Coinchoose, and then multiply it by 1000 for the approx scrypt to SHA256 conversion rate, you will see what I mean. That's still only 10% of the net hash.

GPU's are not a good example, as they chew a lot of power, it's probably more relevant to talk about the 110nm Avalon and ASICMiner chips out there. No ROI on batch #3, Block Erupters etc. but 110nm represents most of the current net hash. There is nothing wrong with the 110nm ASIC tech except it was sold at a ripoff price to early adopters. That stuff could easily be pumped out below $10 per GH/s. to make it competitive again.

More expensive means higher MH/$.  Why would anyone want that.  I even pointed out it is likely there will be more efficient 28nm chips but given the higher cost of 20nm chips a given design (any design of any efficiency) would cost MORE not LESS at 20nm.  Once again why would anyone want MORE expensive chips?

The idea of using low-powered unprofitable equipment to "support" or "protect" the network is flawed. Hardware that is no longer profitable to use for mining will probably not have a significant contribution to the total hash power of the network.

Imagine if all the GPUs started hashing again. They would contribute at most 25 TH/s, or less than 5% of the current total hash power.

Is ROI from just mining important? It seems to be the holy grail.

The way I look at it, is if I mine with gen 1 ASICs at a loss and it now costs me $50 to mine each BTC, then guess what, I just got BTC for less than half the market price! I can trade on that. Obviously miners that just dump are going to loose out, but miners that know how to make profit trading will have an advantage over traders that paid more for their BTC. My ROI and more will come from the trading.

People who would not make any money mining 28nm devices.

Just because you bought 28nm pre-order, it does not mean every ASIC designer out there will not try anything that is more efficient than your pre-order.  Looks like you are trying to convince yourself that your purchase of hashfast will not become obsolete for a very long time.

The ASIC train is moving faster than anyone can say: WTF just happened...

Avalon already said they will be working on Gen 2, 3 and 4 (at the same time?).  Are you ready for your WTF moment?

Avalon, Bitfury, and ASICMiner may launch chips using a smaller process in 2014 but 28nm is cheaper than 20nm.  Who would want to buy MORE expensive chips.  I think people just assume 20nm = cheaper than 28nm (or x < y using any two process nodes).  That isn't true.  Take a look at NVidia's chart for each process there is a date where the smaller process became more cost effective then the older one.  That doesn't occur the day the process is available.   In 2015 20nm will be EXPENSIVE more expensive than the more mature 28nm tech.  The only place where it will make sense is products where power usage & heat are more important than raw computing power.  Can you think of any products made by Apple that might fall into that category?

KNC said "2nd gen" they never said it would be a smaller process.  That is just an assumption being made.  If 20nm is more expensive per transistor and requires NRE why would KNC be launching 20nm tech?  Their customers want to pay more for less hashing power and have to wait for development relative to the highly available (in 2014) 28nm tech?

Avalon & ASCIMiner will likely launch 55nm or better versions in late 2013.  If they don't they might as well get out of the business because they will have no customers by 2014.  Bitfury is kinda an unknown.  For 55nm their tech is pretty friggin good, it has higher efficiency (MH/W and MH/mm^2) then KNC 28nm chip (based on pre-announced specs by both companies). It may simply not be worth the NRE to jump to 28nm if they can compete at 55nm.

I am thinking Avalon via TSMC, and KNCminer gen 2. That's assuming the staff of Avalon haven't been lynched by an organized crime gang or something.

 

I think you are confusing "what is technically possible" with what is ECONOMICAL.   The price per transistor for 20nm is projected to be HIGHER than 28nm through 2014 (and possibly 2015).  That means same design, same efficiency, die shrinked down will be MORE EXPENSIVE (MH/$) than 28nm.  That is before you consider the new multi million dollar 20nm NRE which will have to be ammortized over the initial chips.


Once again since this doesn't seem to sink in (look at NVidia chart).  New tech generally has a HIGHER cost per transistor for a couple years.  That means Avalon dumping millions into 20/16nm would be an idiotic move.  It would allow them to produce chips which cost them more (MH/$) than it cost their 28nm counterparts.

Bitcoin tech will move to 20nm when the cost per transistor is CHEAPER than 28nm.  The bad news (see NVidia slide) is that there have been two trends in the last decade.  The first is that the TIME since process node is available until it is mature enough that is is CHEAPER is taking longer and longer with each process node shrink.  It initially was 1 years then became 2 and now is looking more like 3 to 4.  The second bad news is that even when mature each node is showing a smaller and smaller gain on the prior node.  Look at the difference between 80nm and 55 nm lines around 2011, now compare it to 40nm and 28nm around 2013.  Now look at the future proposed price improvements.

A 20nm chip would have to be significantly cheaper per MH/$ in raw cost so that it can absorb the increased NRE cost and the increased risk.  Just because 20nm is "available" doesn't mean those conditions apply.  2015 is conservative.  It might be 2017 or 2018 before those conditions apply.

Exactly.  I am not saying 20nm is impossible but for it to make sense with Bitcoin it would have to have a lower cost per transistor than 28nm that isn't projected to happen until 2015 or later.  Even when the cost per transistor reaches parity with 28nm you still have the issue or a brand new NRE.  The cost will need to decline substantially so that even with NRE 20nm is cheaper to make than 28nm.   That isn't happening in 2014, maybe 2016.

20nm being "available" means nothing if it can't be priced under 28nm miner tech.  Nobody is going to pay a premium on new miners when ROI% are stretched out to 2+ years due to massive over capacity just to be cutting edge.  Proven, highly available, "good enough", and cheap 28nm will be the name of the game for 2014.

 IMO, while it should be technically possible, gut-check says I don't see it happening for Bitcoin ASICs in 2014 due to excessive NRE costs. If we're extremely lucky, we might hear about something near Christmas 2014, but more than likely 2015 before being able to own physical 20nm ASICs.

For sure we will see some 20nm next year, but will it be better ROI than 28nm? Normally new tech has a premium. There is lots of margin on the current 28nm ASIC offerings, so room to move there.

 

Not sure if you were agreeing or disagreeing but given no Bitcoin ASIC company is on the same scale as Apple and Apple isn't looking at volume 20nm shipments until end of 2014 it kinda shows how far out 20nm is right now.

http://www.electronicsweekly.com/mannerisms/manufacturing/apple-signs-tsmc-for-16nm-and-10nm-nodes-2013-06/

http://www.eetimes.com/document.asp?doc_id=1264668

^^^ scam.  We need a scammer tag.

I am making a 11 nm hashing chip. It does 800GH/s and I am selling them for 200 BTC each. Pre-order now. http://www.Minegoto11.com

The economic brick wall is 28nm.  20nm is beyond rediculously expensive for anyone who doesn't know their own fab (Intel, Samsung).  Nobody is predicting high availability or decent prices of 20nm wagers from any foundry in 2014 either. Nobody (not even Intel) is using 14nm yet, the technical challenges of 10nm haven't even been addresses.   Still even if 20/14/10nm is techincally available what matters in mining is the cost per hash and Bitcoin mining is almost perfectly parallel.  This means a 20nm chip is going to cost MORE than a 28nm one (everything else being the same) until the cost per transistor is lower on 20nm then 28nm.   Usually that takes 2-4 years from process start and it is taking longer and the end benefits are getting smaller (higher NRE, more delays, more complexity, more yield issues, and overall a shrinking gain on the prior process node).  So 20nm SHA-2 ASICs probably won't make sense even in 2015 or 2016 and 14nm isn't even on the map yet (2020?+).



That means for the convieable future 28nm is as good as it gets.  There may be more efficient 28nm designs but that would mean a brand new NRE cost and it would have to be a lot better to make sense (i.e. a company's new chip would need to be so much cheaper (MH/$ and MH/W) that even after a share of the NRE it added it is still a better deal that then existing design).  It remains to be seen if that will happen.  Regardless once everyone is on 28nm and shipping in volume, the market will become highly saturated and mining margins will be a small percentage over electrical cost of the most efficient miners.  Have a less efficient miner, have higher than normal electrical costs, paying for expensive datacenter space expect it to only be a question of how long before your net operating margin goes negative.  Really no different than GPU mining.  Those with excessive electrical costs or poorly though out rigs could only profit marginally when the price/difficulty was high and when it tanked they had to idle or mine at a loss.

With the recent selloff of ASICMiner shares, maybe they are already screaming?

The big farms like ASICminer need to make a profit or their shareholders will scream.

They will be replaced by more energy efficient ASICs or will be run in areas where electricity is dirt cheap.

Once we reach technological brick wall (10nm), the network hash rate will level off and difficulty might actually stay flat or go down.
At that level, mining=electricity & hosting costs + a tiny profit.

All we can now do is use the machines what they are made for, and i personally will run it as long as the electricity usage is being paid by the machine.
As soon as it becomes a loosing money issue i will probably send it to the recycle industry which will try to scrape out as much usefull materials as possible.
I do not say anything about what bitcoin is doing since i still not got a crystal ball
As soon as i can predict what is going to happen and never fail in that i would be the richest man on the planet.
Since i am one of the simple worker slaves, you can fill in how rich i am.... (NOT if you could not guess it allready, we eu slaves are only good for pay taxes)

Too many assumptions surround the mining outlook predictions.

- that difficulty will keep rising exponentially
- that all mining manufacturers will ship products (many all are SHA-2 ASIC only start-ups, I don't think Samsung are producing miners any time soon)
- that BTC price will stick long-term at $80-130 (what if it plunges tomorrow?)

People are getting caught up in their very fixed view of what should happen given current trends, without paying mind to possible severe trend reversals. Not being bearish, what if the price doesn't stop the current linear expansion? What if a high density 3d printed solar cell technology gets announced before December? What if an oil price shock hits the global market '70s style, causing price inflation madness? Mining depends on alot of variables, and assuming that so many will remain the same to support narrow predictions is pretty risky.

https://bitcointalksearch.org/topic/prediction-a-sha256-alt-coin-will-become-as-big-a-success-as-litecoin-1-2yrs-277864

Wont happen, If not big unprofitable, most will run in a hope of price rise

Depends on the ASIC.  ASIC miners relying on dozens of chips per board will probably hang out for a few years under your desks first, get cannibalized for their PSs and a few case parts, and then get trashed.  Some chippers might offer next-gen chippery in a similar package with identical pinout, power specs & footprint, offering to retrofit them to your boards (this will be more difficult than simple board assembly, more like reflow repair).  Maybe one of the ASIC manufacturers is already thinking in terms of socketed PGA chips (HasFast has a chip rendering on their site that looks exactly like Intel CPUs, down to the heat spreader), so users may replace ASICs like they replace CPUs.

Or you can create a collector's market.  Think NEXT Cube

Asic's have a problem if they are no longer usefull on the task they are made for.

These chips are designed to do one task pretty well, so indeed the question is what can be done with them is limited to calculate sha256 security hashes
In fact they where allready used for this task.

The problem is that these chips can not be changed and reprogrammed like fpga.

So what will your expenssive toy be worth after new and faster designs enter the scene

Also i am kinda sure the companies using asic for other tasks do not want our asics, simply because they pay very little for their asics because the ones they use are made in massive numbers.

We pay enormous prices compared to the normal industry designed ones.

I will be expecting kinda 0.0 value after they no longer can be used on the bitcoin scene
Unless you can find an alternative sha256 coin which can be done on the same machines.

 

People will switch them off to save electricity until such time are they are profitable to run again.

Someone will set up a black market distributed computing network for cracking passwords.  I'm sure there's some demand for a few PHash at rock bottom prices.

Someone will create a clone coin which can be mined with current ASICs

will mine unless the loss from electricity costs are more than a couple hundred dollars a month.

They'll be sold on eBay to those who don't do their research properly.

Hobbiests will run their ASIS forever... even when they are producing .0001 BTC/day. PROs will sell off old equipment when they think they can make more return by putting the effort to run the equipment into a new box.

Yes, I have admitted in the past I view this as a hobby.

Either you are irrelevant hobby miner or brand new.

Well - yeah, I can see that to an extent, but at some point, a solar farm will never pay for itself if all its daily intake is being spent on an ASIC farm mining the equivalent of $10/month. As for people with "free" electricity, no landlord's going to stand for them sucking down $1k in electricity per month on a $600/mo lease - so there's a scalability problem there, too.

I think while there may not be "revolutionary" performance improvements in ASICs over previous generations, there will at least be "revolutionary" improvements in the cost for consumers (as we've been seeing since the first ASICs were announced for pre-order) as ASIC manufacturers suddenly face a massive decline in demand and price equipment x% above cost to manufacture instead of x% below what the other guys are charging. That doesn't even factor in manufacturers becoming established/experienced and implementing real manufacturing techniques (instead of half a "line" being a row of plastic tables), and with dedicated purchasing departments.

Profitability is a matter of perspective. Just like how an orebody can be ore one day, and dirt the next (ore is defined as being profitable to mine, once the costs outweigh the profits the rock is no longer considered to be ore), what is unprofitable to some people will be profitable to others.

Consider the requirements for operating mining equipment. You have inputs of power and an output of hashes. Those generally need to be put to a pool and onto the main network, so lets also call bandwidth an input here.

The cost of bandwidth for mining is almost negligible, and in most cases people pay a set amount for their bandwidth whether they use it or not. Power is a much bigger concern, however the cost of this varies from place to place - some pay a fortune for it while others may be able to get it for free.

Take my setup for example, my rigs are solar powered. I have costs of bandwidth but zero on power. If you want to get technical you could potentially factor in the cost of batteries and their associated degradation, but equally you could build a system without batteries, or using any other source of renewable "free" energy. These operations will in essence always be profitable, while those with high power costs will end up unprofitable much sooner.

What is likely to happen is those with highest power cost will end up selling their older equipment, which will be purchased for bargain basement prices by those with no concern for power costs. This will be in part replaced with newer more efficient gear, however the network hashrate won't increase anywhere near the rate it has with the transition from GPU and FPGA to ASICs, as there really isn't a huge improvement to be had.

The second hand market will be decidedly different to that of FPGAs and GPUs. These can both be used for other things, while ASICs can't (unless someone else figures out a use for SHA256(SHA256(x)) hashes)

I can definitely see them tossed if they no longer work, I can even see some tossed if it's too expensive to run them.  ASIC's are not like the old 8088 computers that take 10 days to do what my system can do in 10 minutes.  Unless someone cracks quantum computing, I see a minority being trashed.

I dunno, I'm entertained everytime I get a message in the corner of my screen that I got sent BTC from my pool.  While it doesn't not happen as often as it used to, it does still happen.

What'll happen to ASICs when they're no longer profitable to run?

BFL will then ship them to customers.


(thank you, thank you very much)

Oh my, the things people come up with...

They will end up in dumps like old computer hardware today, what did ya think?


ASIC are a pro tool and not some home entertainment center.

You have different goups of people out there.  Some will sell their equipment, some will discard it, some may try an alt coin (doubtful), the rest will keep mining.  Since the ASIC is built for only 1 purpose, I really cannot see them being repurposed.  I cannot see turning a Formula 1 car into a family sedan.

I do not see how running any type of equipment could be a disservice.  I saw someone post that where they worked they had a server farm built and tested, but the customer would not be 'occupying' it for another month or so, and he decided to point them all at Primcoin to do some earning instead of sitting idle.  The only downside of obsolete equipment as you call it is that you are spending more than you are earning.  At that point it pretty much becomes a hobby.

If the manufacturers actually CAN make something that is better, I am sure they will.  The question will become whether their ability to improve upon what they have now will be profitable when sold to the customer *and* if the customer thinks it is worth buying.  The PS4 is coming out soon, yet I still have a PS2 and a PS3.  Even though I plan on purchasing one, I plan on keeping them, others may or may not agree.  

alt coins are mined as they have lower difficulty and will ultimately be convertable to btc.

I don't see many folks turning off ASIC's in the next year.  I think miners are bulls by nature and so most of us here seem to think price will increase (which obviously doesn't affect BTC ROI but that is a very different discussion).  I think a lot of miners are hoping for an increase so they can stave off saturation point longer! 

Assuming no increase it's kind of hard to say what people will do.  I think there are just too many factors on an individual level for there to be a real majority of people doing one thing.

Factors like:
"free" electricity, hobbyist (helping secure/decentralize network, willing to run at slight loss), capital investment of hardware, efficiency of hardware (130nm or 110nm Avalon vs 28nm everyone else), how much (if any) profit was made before "saturation point" is reached (when electrical cost for 1 BTC = 1 BTC mined).

Because of the imperfect information system I can see us overshooting the mark (by how much I won't guess) and a lot of people will run these machines quite a while mining negligible amounts chasing their investment.
 

I've been wondering this for a while.... Tens of millions of dollars in equipment will lose money if run in roughly a year (assuming price doesn't skyrocket).

So what happens with it? Will organizations, perhaps the Bitcoin Foundation, buy these units and keep the hashpower on the network for added security, running them at a slight loss, perhaps? Will they be repurposed somehow? Will they simply be thrown away? Will the current owners just keep mining?

While on the topic -- is running obsolete equipment a disservice (long-term) to Bitcoin by making newer, more effective ASICs less attractive? People talk about running their GPUs, FPGAs, and ASICs (in the future) for the sake of the network, but isn't that actually just making it less likely for people to buy or design/develop/manufacture new ASICs since the profit projection would be lower?