Takeshi_Kovacs
Member
Offline
Activity: 65
Merit: 10
|
|
September 02, 2013, 05:29:46 PM |
|
My projections of a 41M linear difficulty rise (not exponential!), starting on March 1st with a global hashrate for 20PH/s, the 600GH card will generate 12 BTC (and a few more in all of 2015). I do assume it will be exponential until early Feb 2014, which brings us to 14PH/s at the beginning of Feb..
what dıd ı just read Maybe the spreadsheet is easier to see... https://docs.google.com/spreadsheet/ccc?key=0Auya3iRE6az1dDc0UVgwMU52YVpTazVjSHByOGNiWHc#gid=8The idea is that continuous exponential growth is ridiculous. In like 3 years the solar system's resources are committed to bitcoin mining. So... I made a spreadsheet that goes exponential for just the committed hashrates of the announced vendors which puts out about 14PH/s of global hashrate. Using the current growth rate and mapping it to an exponential is suprisingly close. Here is the wolfram alpha difficulty data mapped starting in Feb through the next projected retarget (which I undershot by a picking 85M): http://www.wolframalpha.com/input/?i=exponential+fit+%7B%7B13%2C+3651012%7D%2C%7B37%2C+4847647%7D%2C%7B96%2C+11187257%7D%2C+%7B120%2C+15605633%7D%2C%7B156%2C26162876%7D%2C%7B167%2C31256960%7D%2C%7B189%2C50810340%7D%2C%7B200%2C+65750060%7D%2C%7B210%2C85000000%7D%7DI figure that rate will continue until all the pre-orders are shipped, which puts the global hash rate at 14PH/s in early Feb. At that point I assume the growth rate will be linear not exponential (i.e. the growth between retargets will be constant rather than increasing... i.e. no longer accelerating in difficulty). I basically thought that at that point, the rate would "coast" at that linear rate... rather than being pushed. The curve flattens out with a slope equal to the "exit velocity" of the exponential portion... around 41Million daily difficulty rise. We are currently at around 1.6M daily difficulty rise, so this linear phase I am "hoping" is only 25-ish times larger than what we are seeing today. Anyway, if you use the spreadsheet I posted, with these assumptions, put in a delivery date of March, 1st (an optimistic BFL target date), 600 GH performance, and check out the cumulative BTC mining performance it comes out to 12 BTC in 2014, and 5BTC in 2015 (though projecting 2014 is fraught enough, 2015 is even more fantasy). Obviously, this could (and most likely is) complete bullcrap... reality may not be anything like this projection. But modelling is fun, and it puts in high relief delays in delivery are going to shorten these miners shelf lives considerably. Even the fancy 28nm ones. Hope that helps. Edit: some grammar/spelling changes. my braın ısnt workıng well today. what the hell ıs a ph? ıs that supposed to be hıgher then a thPeta is 1000 up on tera. http://en.wikipedia.org/wiki/Peta-
|
|
|
|
Tomintx
Member
Offline
Activity: 69
Merit: 10
|
|
September 02, 2013, 10:20:58 PM |
|
Fun with projections: A 100GH/s unit delivered today will never earn beyond $3K in its lifetime. Given the current exponential growth of difficulty, A 100GH/s unit will earn $3K by Feb 2014, and then earn another $20 by Jan 1, 2015. A new 600GH unit from BFL, delivered in January (I know, I'm smoking something, but so are they!) will earn $675 in a year. $665 of that by June 1, 2014. Not much payback on a $5K "investment." In 2015, it would not even be earning a penny a week. Current growth rate (26% per 11 days) puts difficulty at 1B+ on Jan 1, 2014. Current growth rate (26% per 11 days) puts difficulty at 2.8T+ on Jan 1, 2015. If all the GPU miners in the WORLD shut down tomorrow, it won't affect things much. What was the total hash rate before the Avalons, and ASIC-miners (and BFL's) started shipping? It actually looks like the success of BTC mining is also going to be the downfall of BTC mining. Growth makes mining unprofitable...
|
|
|
|
Cluster2k
Legendary
Offline
Activity: 1692
Merit: 1018
|
|
September 03, 2013, 12:04:20 AM |
|
Don't forget to add 10% GST plus Customs duty for processing the paperwork to get the unit into Australia.
|
|
|
|
bcp19
|
|
September 03, 2013, 12:07:36 AM |
|
Fun with projections: A 100GH/s unit delivered today will never earn beyond $3K in its lifetime. Given the current exponential growth of difficulty, A 100GH/s unit will earn $3K by Feb 2014, and then earn another $20 by Jan 1, 2015. A new 600GH unit from BFL, delivered in January (I know, I'm smoking something, but so are they!) will earn $675 in a year. $665 of that by June 1, 2014. Not much payback on a $5K "investment." In 2015, it would not even be earning a penny a week. Current growth rate (26% per 11 days) puts difficulty at 1B+ on Jan 1, 2014. Current growth rate (26% per 11 days) puts difficulty at 2.8T+ on Jan 1, 2015. If all the GPU miners in the WORLD shut down tomorrow, it won't affect things much. What was the total hash rate before the Avalons, and ASIC-miners (and BFL's) started shipping? It actually looks like the success of BTC mining is also going to be the downfall of BTC mining. Growth makes mining unprofitable... Difficulty will never reach 2.8T by 2015. 1 Petahash will equal 140 million 1000 petahash would then be 140 billion. You'd need 20,000 petahash hit 2.8T difficulty. That would equal 33,333,333 Monarchs.
|
I do not suffer fools gladly... "Captain! We're surrounded!" I embrace my inner Kool-Aid.
|
|
|
Tomintx
Member
Offline
Activity: 69
Merit: 10
|
|
September 03, 2013, 01:49:02 AM |
|
Difficulty will never reach 2.8T by 2015. 1 Petahash will equal 140 million 1000 petahash would then be 140 billion. You'd need 20,000 petahash hit 2.8T difficulty. That would equal 33,333,333 Monarchs.
I don't quibble with your math. I don't think that 33M Monarchs will be delivered by 1/1/2015. I would question whether or not ONE Monarch would be delivered by 1/1/2015. However BFL are not the only ones producing hashing chips. The hashing algorithm/circuitry is very conducive to massive parallelism, and the design is not all that complicated. Anyone with access to the tools can create a design, and once you have the GDSII, you can go to just about any fab with it. The fabs in China are well entrenched at the 28nm node, and will be producing either 20nm or 16nm before long. I think that we're going to see continued escalation in the mining war, and the PHash/sec milestone is not that far off. Current growth in the hashing rate has been 17% per 11.4 days so far this year. If we extend that rate into the future, we'll hit the 1PH/sec milestone in December 2014. Using these numbers, your $4680 Monarch, if it goes online on 1/1/14, will earn 27.4 BTC before leveling off at < .01 BTC/week in November 2014. You'll have earned 27 BTC by July, not so much after that. I'm glad I'm not waiting for a $30K "minirig" to show up. If it got here today, I'd earn $80K. If it gets here on December 1st, I'd probably get half my money back. I'd certainly trade it for 6 or more Monarchs. I'd be lucky to break even. I'm still waiting for a single Jally, which might get here by 10/1. I'll be lucky to break even.
|
|
|
|
jspielberg
|
|
September 03, 2013, 01:56:17 AM |
|
It doesn't look that good even if the difficulty only goes linear at 41M daily difficulty rise (15B difficulty at the end of 2014). With the best case scenario of them shipping out the cards at the end of 2013, they will end up makeing 21 BTC in 2013 and then 3 or 4 more in all of 2015. I would say these projections are pretty optimistic being that they aren't exponential but linear. https://docs.google.com/spreadsheet/ccc?key=0Auya3iRE6az1dDc0UVgwMU52YVpTazVjSHByOGNiWHc#gid=8Basically... 600 is too small for delivery in December (and given BFL's track record, 2013 delivery is laughable)
|
|
|
|
bcp19
|
|
September 03, 2013, 02:55:00 AM |
|
Difficulty will never reach 2.8T by 2015. 1 Petahash will equal 140 million 1000 petahash would then be 140 billion. You'd need 20,000 petahash hit 2.8T difficulty. That would equal 33,333,333 Monarchs.
I don't quibble with your math. I don't think that 33M Monarchs will be delivered by 1/1/2015. I would question whether or not ONE Monarch would be delivered by 1/1/2015. However BFL are not the only ones producing hashing chips. The hashing algorithm/circuitry is very conducive to massive parallelism, and the design is not all that complicated. Anyone with access to the tools can create a design, and once you have the GDSII, you can go to just about any fab with it. The fabs in China are well entrenched at the 28nm node, and will be producing either 20nm or 16nm before long. I think that we're going to see continued escalation in the mining war, and the PHash/sec milestone is not that far off. Current growth in the hashing rate has been 17% per 11.4 days so far this year. If we extend that rate into the future, we'll hit the 1PH/sec milestone in December 2014. Using these numbers, your $4680 Monarch, if it goes online on 1/1/14, will earn 27.4 BTC before leveling off at < .01 BTC/week in November 2014. You'll have earned 27 BTC by July, not so much after that. I'm glad I'm not waiting for a $30K "minirig" to show up. If it got here today, I'd earn $80K. If it gets here on December 1st, I'd probably get half my money back. I'd certainly trade it for 6 or more Monarchs. I'd be lucky to break even. I'm still waiting for a single Jally, which might get here by 10/1. I'll be lucky to break even. I used the monarch as it's about the biggest thing out there. 33 million monarch would equal $150 billion. Even if they came down to $100 that'd be $3.3 billion. Even at $10 I can't see people investing $330 million in the next year, not to mention the unlikelyhood of $13/TH. 10-14PH looks possible next year, 1.4 to 2.0 billion difficulty. Anything above that is kinda unrealistic.
|
I do not suffer fools gladly... "Captain! We're surrounded!" I embrace my inner Kool-Aid.
|
|
|
Dorrace (OP)
Member
Offline
Activity: 70
Merit: 10
|
|
September 03, 2013, 10:42:55 AM |
|
Precisely but 41m a day the difficulty has been the same for 2 weeks it seems to me the sudden burst was due to large deliveries from bfl and a couple other companies I think we're safe for a few weeks more
|
|
|
|
Dorrace (OP)
Member
Offline
Activity: 70
Merit: 10
|
|
September 03, 2013, 11:14:26 AM |
|
also I think 100,000 ASIC miners is more realistic, 40-50% being monarchs over the next 12 months
Average it out at 500gh/s that is 5ph/s in the next year. Isn't it?
Yes there has been substantial growth in difficulty but I don't believe they will bring out a new ASIC in less then 12 months it defies market logic. Even the monarch has been pushing it but they know the monarch won't be ready for another 5-6 months .
Speaking from a marketing point of view they would know that the ship date is a bit early but worst case scenario they will ship 2 months late. So 6 months til you get your batch of monarch. February -march well see the difficulty jump again. Around January December well see the full range of 4th generation miners from bfl. They will have a 4month minimum ship date. Best case scenario 90 days if bfl ups its shit and doesn't have the delays we saw this year.
Bfl would have folded by now if it planned on running off with out money and it wouldn't have made such a bad start to its business by pissing off customers. It's not too good to be true because it isn't too good we have to wait months for our ASics they aren't scamming us people have received their units ordered in the prior year.
Now realistically speaking lets make a realistic chart. You can't go and make a chart based on recent growth because there is simply no way the worlds ASIC companies will be able to manufacture 33,000,000 ASIC miners capable of 500gh/s minimum outputs.
I'm going to put together some data and see what we come up with I suggest some of you do the same.
Taking into account that list of the ASIC miners ordered from bfl and we will have to assume it makes up at least 50% of the recent hash rate increase.(I'm not quite sure what percent of the market they hold).
Then we take into account how much they have produced and shipped in the last 14 months work out their production rate add a reasonable percent , 30-35% increased production.
I can't be bothered looking it up right now I'm on my phone but when I get infrint of a pc ill source some data. Ill put up a poll later to get rough numbers of ASIC companies customers etc.
There is just too much horseshit speculation going on and I don't like it. I honestly believe you can still turn a profit in 12 months from an ASIC unit. We should also work out an equation of hash rate to difficulty to use as a guideline when considering ASIC units. If there already is one please help and show us. Not looking at usd value though try not to use fiat currency as a guide when looking at profit , ROI . Lets use btc.because that minimises exchange rate speculation. We also round up costs and round down returns this provides a worst case scenario.
For starters we have 100,000 500gh/s units =5ph/s added to our current hash rate we can round that up to 6ph/s from memory the current hash rate was above 500th/s?
Can someone calculate expected difficulty at 6ph/s?
But until we get realistic numbers of ASIC units preordered in the 500gh category this data will also be speculation but what I believe to be solid speculative data. A lot more trustworthy then a lot of the speculation we have been getting / making.
Assuming difficulty is 75,000,000 assuming network hash rate is around 650-750th/s Ill roughly say our hash rate with 6ph will be around 650,000,000-750,000,000 does anyone agree? Around 1,000,000,000 diff this time next year? That will return something like 0.20-0.30 btc a day If I'm not wrong
Please someone add to this or adjust . I want as much input as we can get.
When you use a mining calculator assume electricity costs 0.35 cents per kwh (worst case scenario we pay around this in Australia).
Simply we cannot assume the difficulty will rise at a steady rate , ASIC miners do not go up in hash rate the longer they are plugged in. Difficulty and hash rate is directly manipulated by how much a company can produce and sell ASIC units.
|
|
|
|
bcp19
|
|
September 03, 2013, 11:20:03 AM |
|
It doesn't look that good even if the difficulty only goes linear at 41M daily difficulty rise (15B difficulty at the end of 2014). With the best case scenario of them shipping out the cards at the end of 2013, they will end up makeing 21 BTC in 2013 and then 3 or 4 more in all of 2015. I would say these projections are pretty optimistic being that they aren't exponential but linear. https://docs.google.com/spreadsheet/ccc?key=0Auya3iRE6az1dDc0UVgwMU52YVpTazVjSHByOGNiWHc#gid=8Basically... 600 is too small for delivery in December (and given BFL's track record, 2013 delivery is laughable) You chart looks nice and all, but since the difficulty only retargets every 2016 blocks, your daily diminishing returns paint a far bleaker picture than reality. In order to get your 41M daily increase, there would have to be more than 300TH added to the network every day. That is the equivalent of 600 mini-rig units, or 500 monarchs or 150 TerraMiner IV's. You are talking $2,100,000-3,000,000 DAILY being spent on equipment and that's using the best $/TH ratios currently out there. Over a quarter of a billion dollars by the end of the year and another billion next year. You'd be closer to reality if you factored in that 41M on a per week basis.
|
I do not suffer fools gladly... "Captain! We're surrounded!" I embrace my inner Kool-Aid.
|
|
|
bcp19
|
|
September 03, 2013, 11:38:12 AM |
|
also I think 100,000 ASIC miners is more realistic, 40-50% being monarchs over the next 12 months
Average it out at 500gh/s that is 5ph/s in the next year. Isn't it?
Yes there has been substantial growth in difficulty but I don't believe they will bring out a new ASIC in less then 12 months it defies market logic. Even the monarch has been pushing it but they know the monarch won't be ready for another 5-6 months .
Speaking from a marketing point of view they would know that the ship date is a bit early but worst case scenario they will ship 2 months late. So 6 months til you get your batch of monarch. February -march well see the difficulty jump again. Around January December well see the full range of 4th generation miners from bfl. They will have a 4month minimum ship date. Best case scenario 90 days if bfl ups its shit and doesn't have the delays we saw this year.
Bfl would have folded by now if it planned on running off with out money and it wouldn't have made such a bad start to its business by pissing off customers. It's not too good to be true because it isn't too good we have to wait months for our ASics they aren't scamming us people have received their units ordered in the prior year.
Now realistically speaking lets make a realistic chart. You can't go and make a chart based on recent growth because there is simply no way the worlds ASIC companies will be able to manufacture 33,000,000 ASIC miners capable of 500gh/s minimum outputs.
I'm going to put together some data and see what we come up with I suggest some of you do the same.
Taking into account that list of the ASIC miners ordered from bfl and we will have to assume it makes up at least 50% of the recent hash rate increase.(I'm not quite sure what percent of the market they hold).
Then we take into account how much they have produced and shipped in the last 14 months work out their production rate add a reasonable percent , 30-35% increased production.
I can't be bothered looking it up right now I'm on my phone but when I get infrint of a pc ill source some data. Ill put up a poll later to get rough numbers of ASIC companies customers etc.
There is just too much horseshit speculation going on and I don't like it. I honestly believe you can still turn a profit in 12 months from an ASIC unit. We should also work out an equation of hash rate to difficulty to use as a guideline when considering ASIC units. If there already is one please help and show us. Not looking at usd value though try not to use fiat currency as a guide when looking at profit , ROI . Lets use btc.because that minimises exchange rate speculation. We also round up costs and round down returns this provides a worst case scenario.
For starters we have 100,000 500gh/s units =5ph/s added to our current hash rate we can round that up to 6ph/s from memory the current hash rate was above 500th/s?
Can someone calculate expected difficulty at 6ph/s?
But until we get realistic numbers of ASIC units preordered in the 500gh category this data will also be speculation but what I believe to be solid speculative data. A lot more trustworthy then a lot of the speculation we have been getting / making.
Assuming difficulty is 75,000,000 assuming network hash rate is around 650-750th/s Ill roughly say our hash rate with 6ph will be around 650,000,000-750,000,000 does anyone agree? Around 1,000,000,000 diff this time next year? That will return something like 0.20-0.30 btc a day If I'm not wrong
Please someone add to this or adjust . I want as much input as we can get.
When you use a mining calculator assume electricity costs 0.35 cents per kwh (worst case scenario we pay around this in Australia).
Simply we cannot assume the difficulty will rise at a steady rate , ASIC miners do not go up in hash rate the longer they are plugged in. Difficulty and hash rate is directly manipulated by how much a company can produce and sell ASIC units.
6PH would be ~825M difficulty. While I agree BFL would have already run if that was their intention, giving them 50% of the current increase is bit generous in my books. I'd estimate closer to 25-33%. Bitfury added 125TH as seen at ghash.io and has shipped their starter kits. ASICMiner has a lot of quantity but little actual network growth. Avalon's chips are in the wild, so they are still an unknown right now. There are several companies poised to flood the market, so the 500TH gain of last month could easily be anywhere from 1,000-2,000TH. The problem with these calculators is they look at % growth which quickly gets out of hand when using today's numbers. 500 last month becoming 1,000 this month is feasible, 2,000 next month is possible, 4,000 the month after shaky, 8,000.. 16,000... 32,000.. 64,000... maybe if you got Intel and AMD into making ASICs. 6PH by year end can happen IF all of the current companies ship on time and fill all pre-orders. Kind of a big IF from the current viewpoint.
|
I do not suffer fools gladly... "Captain! We're surrounded!" I embrace my inner Kool-Aid.
|
|
|
jspielberg
|
|
September 03, 2013, 01:20:14 PM |
|
It doesn't look that good even if the difficulty only goes linear at 41M daily difficulty rise (15B difficulty at the end of 2014). With the best case scenario of them shipping out the cards at the end of 2013, they will end up makeing 21 BTC in 2013 and then 3 or 4 more in all of 2015. I would say these projections are pretty optimistic being that they aren't exponential but linear. https://docs.google.com/spreadsheet/ccc?key=0Auya3iRE6az1dDc0UVgwMU52YVpTazVjSHByOGNiWHc#gid=8Basically... 600 is too small for delivery in December (and given BFL's track record, 2013 delivery is laughable) You chart looks nice and all, but since the difficulty only retargets every 2016 blocks, your daily diminishing returns paint a far bleaker picture than reality. In order to get your 41M daily increase, there would have to be more than 300TH added to the network every day. That is the equivalent of 600 mini-rig units, or 500 monarchs or 150 TerraMiner IV's. You are talking $2,100,000-3,000,000 DAILY being spent on equipment and that's using the best $/TH ratios currently out there. Over a quarter of a billion dollars by the end of the year and another billion next year. You'd be closer to reality if you factored in that 41M on a per week basis. I think you are making the assumption that all that hashpower will be retail. I would venture to say the endgame is getting close (middle of 2014?). Asic Producers will be self mining as it won't be profitable to sell low margin chips when they can just self-mine at $.5/GH ($500/TH). This assumes they have been through a pre-order retail round for NRE recpature. I believe that level is sustainable by the different producers who want to keep their 10 to 15% slice of the network.
|
|
|
|
Dorrace (OP)
Member
Offline
Activity: 70
Merit: 10
|
|
September 03, 2013, 01:46:18 PM |
|
All in all, what is being said is its still going to be somewhat profitable to mine just not a get rich quick scheme anymore.
|
|
|
|
bcp19
|
|
September 03, 2013, 07:05:49 PM |
|
It doesn't look that good even if the difficulty only goes linear at 41M daily difficulty rise (15B difficulty at the end of 2014). With the best case scenario of them shipping out the cards at the end of 2013, they will end up makeing 21 BTC in 2013 and then 3 or 4 more in all of 2015. I would say these projections are pretty optimistic being that they aren't exponential but linear. https://docs.google.com/spreadsheet/ccc?key=0Auya3iRE6az1dDc0UVgwMU52YVpTazVjSHByOGNiWHc#gid=8Basically... 600 is too small for delivery in December (and given BFL's track record, 2013 delivery is laughable) You chart looks nice and all, but since the difficulty only retargets every 2016 blocks, your daily diminishing returns paint a far bleaker picture than reality. In order to get your 41M daily increase, there would have to be more than 300TH added to the network every day. That is the equivalent of 600 mini-rig units, or 500 monarchs or 150 TerraMiner IV's. You are talking $2,100,000-3,000,000 DAILY being spent on equipment and that's using the best $/TH ratios currently out there. Over a quarter of a billion dollars by the end of the year and another billion next year. You'd be closer to reality if you factored in that 41M on a per week basis. I think you are making the assumption that all that hashpower will be retail. I would venture to say the endgame is getting close (middle of 2014?). Asic Producers will be self mining as it won't be profitable to sell low margin chips when they can just self-mine at $.5/GH ($500/TH). This assumes they have been through a pre-order retail round for NRE recpature. I believe that level is sustainable by the different producers who want to keep their 10 to 15% slice of the network. Never happen. BTC is too volatile a price point and power costs would be sky high. Companies rely on Fiat, they'd have to sell daily or risk losing value. Imagine 10-15% of all bitcoins mined being sold every single day. Who'd want to buy them knowing it's coming from these greedy companies who refuse to sell the equipment and mine with it instead?
|
I do not suffer fools gladly... "Captain! We're surrounded!" I embrace my inner Kool-Aid.
|
|
|
k9quaint
Legendary
Offline
Activity: 1190
Merit: 1000
|
|
September 03, 2013, 08:39:35 PM |
|
Tips for calculating the return of hardware:
Don't extrapolate out more than 6 months. There are simply too many unknowns and hash rate is growing too fast during a disruptive period to make predictions past that point. If the delivery date is more than 6 months away, don't bother buying it because you cannot anticipate what sort of conditions you will receive the hardware under.
Don't include exchange rate in your calculations.
|
Bitcoin is backed by the full faith and credit of YouTube comments.
|
|
|
thy
|
|
September 05, 2013, 03:41:07 PM |
|
Profitability of the new bfl monarchs (600ghs)
I would say it's looking to be a certain EV- investement even if they would deliver on time, there's no chance you will make your btc's back on an investment in BLF's ASIC at the prices they charge now.
|
|
|
|
Dorrace (OP)
Member
Offline
Activity: 70
Merit: 10
|
|
September 05, 2013, 03:58:33 PM |
|
Profitability of the new bfl monarchs (600ghs)
I would say it's looking to be a certain EV- investement even if they would deliver on time, there's no chance you will make your btc's back on an investment in BLF's ASIC at the prices they charge now. ı admıt the current prıces are expensıve because they dont delıver for atleast 5 months . by the tıme ıt ıs ready and shıpped ıts value ıs dropped by atleast 30 percent.
|
|
|
|
Nemo1024
Legendary
Offline
Activity: 1680
Merit: 1014
|
|
September 09, 2013, 01:27:29 PM |
|
At a glance, I read this thread's title as " Probability of the new bfl monarchs (600ghs)"
|
“Dark times lie ahead of us and there will be a time when we must choose between what is easy and what is right.” “We are only as strong as we are united, as weak as we are divided.” “It is important to fight and fight again, and keep fighting, for only then can evil be kept at bay, though never quite eradicated.”
|
|
|
jamesc760
|
|
September 11, 2013, 12:39:18 AM |
|
LOL
"profitability" and "bfl" are mutually exclusive, if you are pre-ordering today. BFL will not deliver your precious monarch by next February, if you order one today; it will be more like December 2014, if then.
|
|
|
|
mgio
|
|
September 11, 2013, 08:40:21 PM |
|
I estimate profitability from buying a BFL Monarch at 0%.
Margin of error of +/- 4%
|
|
|
|
|