Bitcoin Forum
June 23, 2024, 03:32:30 PM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
  Home Help Search Login Register More  
  Show Posts
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 [29] 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 ... 108 »
561  Bitcoin / Hardware / Re: Best MotherBoard for Butterfly Labs Monarch PCI card? on: August 22, 2013, 04:06:17 PM
Here is a fun experiment.   Put 3x 7990s in a sealed 3U or 4U rackmount chassis and try to keep the cards from burning up.  There is a reason that people use open rigs.  Trying to remove that kind of heat is next to impossible.  Even if BFL delivers the form factor is next to useless because you are never going to be able to fill a rackmount server with 3 to 5 of them.

Correct, removing 1-1.7kW of heat generation in a small sealed form factor is challenging.  Water cooling is an option but with the pumps and tubes, it would be quite difficult to fit into a 4U case.  Open air cooling is by far the most cost effective way to deal with this heat like D&T said. 
562  Bitcoin / Hardware / Re: BFL 600GH "Monarch" already postponed to February! on: August 22, 2013, 03:51:44 AM
That does change a lot.  3-4 months in Bitcoin mining is an eternity at this stage.
563  Bitcoin / Bitcoin Discussion / Re: Luke Jr for Boardmembership on: August 21, 2013, 05:36:48 PM
Does the Bitcoin Foundation have its own discussion board?
yes

All memberships include one vote in their membership class and access to our member forums where you’ll be able to provide input on foundation initiatives, join a committee, or volunteer. Industry memberships include additional benefits outlined here.

Ahh. It is strange that their members are over here talking about their organization's matters.

Why would that be strange?   This is the online epicenter of Bitcoin discussion at the moment.  Seems quite normal to me.
564  Bitcoin / Hardware / Re: HASHFAST warranty just TEN DAYS after delivery? on: August 21, 2013, 05:18:22 PM
Avalon has zero warranty.  Did it stop anyone then?  Maybe a few...

Nope, soldout quick. 
565  Bitcoin / Mining / Re: Difficulty Increase Insurance on: August 21, 2013, 04:27:16 AM
Well when the difficulty chart is only heading one direction, who would take the other side of the deal.  If you find someone, I will buy immediately.
566  Bitcoin / Mining speculation / Re: Why is the demand still so high for Miners with returns like these? on: August 21, 2013, 04:12:04 AM
I was just wondering with the increasing difficulty why the demand is still there for some seemingly high priced (though somewhat readily available) miners that don't seem like they can break even.

Because emotion and greed appear to overpower rational thought and impair basic math skills.

Bingo and now collectively we will see what happens and pay that price.  I will repeat it again, we are in a zero-sum game inside of an ASIC arms-race.
567  Economy / Service Announcements / Re: [Accepting Hardware] ASIC Miner Hosting (Managed) - Electricity @0.02/Kwh on: August 21, 2013, 03:06:40 AM
a lot of bitcoiners in the area of wa - cheap electricity in certain areas of washington.
it's like mini silicone valley -

Yep.  It is the future.  Demand has been awesome.  Signed two more customers up today.   With this environment, miners will need all the help they can get.   
568  Bitcoin / Mining speculation / Re: The ceiling of difficulty with 28nm technology on: August 21, 2013, 02:47:24 AM
In this system you want to be under the cap because you are not forced to buy a contract.  I understand just it doesn't apply to hosting.  I wouldn't have interpretable power contracts for a data-center.  The power rate does go down even more if your under the power contract but the discount is not attractive enough to compel me to operate under those terms.  If you're are running a super heavy power load business that can handle small cut-backs, I can see that being a good thing. 

Companies choose to operate on the interruptable schedule in exchange for lower rates. 

There is always a contract involved.  Your choice is interruptable schedule for a lower rate OR non-interruptable schedule for a higher rate.
Why would you pay more for power on Bitcoin mining then you have to?

Of course this is mostly academic it would take a very long Bitcoin operation to even be given the choice.

Quote
I understand just it doesn't apply to hosting.
A datacenters will CHOOSE to pay MORE for power in return not having an interrupt provision.
A aluminum smelter (or Bitcoin farm) will CHOOSE to pay LESS for power in return for having an interrupt provision.

Interesting.  I know three operations in my area and they are all in the "pay the standard rate under the cap and not have interruptions" camp. 
569  Bitcoin / Mining speculation / Re: Will kWh/hash/sec ever go low enough for mass decentralized mining? on: August 21, 2013, 02:45:26 AM
The way that I see it, the amount of energy it takes to perform a hash per second on a general purpose device (phone, pad, netbook, etc.) is more important than the hash rate of the latest miners.  If the first number becomes very low, then perhaps billions of users could make Bitcoin essentially unbreakable and make a modest amount on the side (when billions of people are using Bitcoin, even a few satoshi a week would be nothing to spit at).  I would much rather see mass decentralization than a scenario wherein Bitcoin is kept afloat by a few hundred (or fewer?) mining centers. 

Centralization will continue until on the Bitcoins are mined.  Better question is what will happen once that occurs.
570  Bitcoin / Mining speculation / Re: The ceiling of difficulty with 28nm technology on: August 20, 2013, 09:23:12 PM
Yep, there is an aluminum smelter in the area.   Thanks for the info, I did read through some of the power contract information and some of the items you mention are listed.  

Now imagine you were building a massive hashing farm.  Would you be willing to trade (hypothetically) up to 1% annual outage for 30% lower power rate?  Like the aluminum smelter it doesn't matter if you are producing Bitcoins in any given second just how much you can produce in a year and at what cost.

I would love a contract like that on a smaller scale but power companies generally only make contracts like that with the largest of users because if they need to cut 1% of demand it is 1% of demand not necessarily 1% of customers.  If they need to shed 1MW of demand it is easier to disconnect (under contract) one 1 MW customer then it is to disconnect one hundred 10KW customers.

My strategy is to build a few separate locations to keep under the cap.  

I think you are still misunderstanding. You wouldn't want to be under the cap.  Would you be upset to have cheaper power?  
This is something companies contact for BECAUSE they want very cheap power rates.  Cheaper than the rates offered to customers under non-interruptable contracts.

In this system you want to be under the cap because you are not forced to buy a contract.  I understand just it doesn't apply to hosting.  I wouldn't have interpretable power contracts for a data-center.  The power rate does go down even more if your under the power contract but the discount is not attractive enough to compel me to operate under those terms.  If you're are running a super heavy power load business that can handle small cut-backs, I can see that being a good thing.  
571  Bitcoin / Hardware / Re: HASHFAST warranty just TEN DAYS after delivery? on: August 20, 2013, 09:01:28 PM
30 days? How about 12 months, if you're so certain about your products? Smiley

30 days seems really short for a $4500 computer product - don't most warranties on computer equipment last at least a year??  Even if they offered to sell an extended warranty it seems like a better idea then putting it at 10 days!  Seriously - if I had it for 11 days and it started on fire and they told me they wouldn't cover it (even though I was operating in all the "appropriate" conditions) I'd be seriously pissed off!  Thanks for catching this...

I think we should be a little more reasonable.  This is for the most part, experimental equipment that is going to be hacked and over-clocked by the nature of the Bitcoin network.  

Imho, one year is too long and I would say 10 days seems a tad short.  30-60 days is what I would expect.  When you get a one year warranty, you have to remember that the hardware was produced on a billion dollar production line.  Yes the chips are made on this type of line but the rest of the parts are mostly custom and hand produced before shipping.  

Likely this situation will only get worse as a dwindling supply of Bitcoins will be chase by more and more high powered miners.


-D

572  Bitcoin / Mining speculation / Re: The ceiling of difficulty with 28nm technology on: August 20, 2013, 08:14:52 PM
Yep, there is an aluminum smelter in the area.   Thanks for the info, I did read through some of the power contract information and some of the items you mention are listed. 

Now imagine you were building a massive hashing farm.  Would you be willing to trade (hypothetically) up to 1% annual outage for 30% lower power rate?  Like the aluminum smelter it doesn't matter if you are producing Bitcoins in any given second just how much you can produce in a year and at what cost.

I would love a contract like that on a smaller scale but power companies generally only make contracts like that with the largest of users because if they need to cut 1% of demand it is 1% of demand not necessarily 1% of customers.  If they need to shed 1MW of demand it is easier to disconnect (under contract) one 1 MW customer then it is to disconnect one hundred 10KW customers.

My strategy is to build a few separate locations to keep under the cap. 
573  Bitcoin / Mining speculation / Re: The ceiling of difficulty with 28nm technology on: August 20, 2013, 07:22:04 PM
Yes, you are correct.  I should of been more clear.  My utility has an agreement so local customers get guaranteed access for local power consumption. 

I see the confusing, he was referring to an OPTIONAL contract not the power company just cutting people off.  There are likely customers in your area who do the same thing.  Just using some made up number (large power delivery if often more complex then a fixed kWh rate).  A big power user (think aluminum smelter) might be offered a guaranteed delivery rate of $0.05 per kWh OR they could agree to no more than 1% annual outage and get $0.035 per kWh.  The customer chooses (isn't forced) to accept a possibility of up to 1% downtime in return for a 30% reduction in electrical rates.  For a business like aluminum smelting where you production cost is mostly electricity this is a beneficial arrangement for BOTH the customer and the power company.






Yep, there is an aluminum smelter in the area.   Thanks for the info, I did read through some of the power contract information and some of the items you mention are listed. 
574  Bitcoin / Mining speculation / Re: The ceiling of difficulty with 28nm technology on: August 20, 2013, 07:20:13 PM
Even with other fees, it's well under $0.05/kwh in certain areas.

As for power interruptions, high power consumption sites (industrial loads) can get substantially cheaper power in some areas if they agree to load shed a few times a year (typically for minutes to hours) and basically turn off the main power consumption items to reduce grid load.  You have to be sucking a LOT of power to do this, but in exchange for the power company being able to turn you off with little to no warning, you can get a much reduced power rate.  Bitcoin mining being interrupted doesn't affect it other than the loss of mining capacity during those times, so allowing it to be interrupted and restarted for a reduced power rate year round is likely to be a net win as well.

When you have Hydro-electric power you don't deal with these interruptions. 

What does hydro have to do with it? The interruptions come at times of unusually high load. Typically in mid winter when there will be high loads for heating and lighting then, during the morning peak demand (cooking breakfast and heating water for coffee and showers) or in the early evening peak (cooking dinner or the whole country turning on the kettle when there is an ad break on TV) the network may simply not have the capacity to deliver the full demand and so the high consumption customers who have agreed to interruption at such periods will be turned off. In hotter places, the peak demand will come on summer afternoons as all the air-con systems work flat out.

I have worked in the electricity supply industry in countries that have hydro power and they have contracts that allow power companies to interrupt the supply to large users.

I am not sure how other areas or countries handle it.  We have public owned hydro electric dams in Eastern Washington.   With that ownership, it has some benefits for local power customers.  If you use over 1MW per month you need to purchase power contracts, there may be clauses in there but with the amount of data-centers located out here, I would guess that it is not just because of the cheap power but also because of the priority access they can have to power.   I talked with my utility quite extensively about this and they told me that there was no issue about peak demand and furloughing customers. I too am still learning about this industry.
575  Bitcoin / Mining speculation / Re: The ceiling of difficulty with 28nm technology on: August 20, 2013, 07:10:54 PM
Even with other fees, it's well under $0.05/kwh in certain areas.

As for power interruptions, high power consumption sites (industrial loads) can get substantially cheaper power in some areas if they agree to load shed a few times a year (typically for minutes to hours) and basically turn off the main power consumption items to reduce grid load.  You have to be sucking a LOT of power to do this, but in exchange for the power company being able to turn you off with little to no warning, you can get a much reduced power rate.  Bitcoin mining being interrupted doesn't affect it other than the loss of mining capacity during those times, so allowing it to be interrupted and restarted for a reduced power rate year round is likely to be a net win as well.

When you have Hydro-electric power you don't deal with these interruptions. 

All power grids have to deal with DEMAND > CAPACITY.
You either
a) use very expensive (in terms of cost per KW/h) "peaking plants" normally natural gas turbines because they can rapidly adjust load
b) you offer your largest customers a deal where they have to idle/scale back (they often do this because b is cheaper than a)
c) you have brownouts and destroy electronic equipment on a masive scale
d) you have rolling blackouts

A hydro-electric dam has a max capacity.  No power grid runs ONLY on hydro-electric dams and even if they did if the total instantaneous demand is greater than the total power grid capacity then something has to give.  Power companies have found that contracting to lower the load on large users is often the cheapest solution.   If 99% of the time your customers consume <10 GW of power and the other 1% of the time it can spike to 11 GW why build out an extra 1GW of capacity which is only used 1% of the time.  The use of hydro electric power doesn't change that dynamic.


Yes, you are correct.  I should of been more clear.  My utility has an agreement so local customers get guaranteed access for local power consumption. 
576  Bitcoin / Mining speculation / Re: The ceiling of difficulty with 28nm technology on: August 20, 2013, 06:49:44 PM
Even with other fees, it's well under $0.05/kwh in certain areas.

As for power interruptions, high power consumption sites (industrial loads) can get substantially cheaper power in some areas if they agree to load shed a few times a year (typically for minutes to hours) and basically turn off the main power consumption items to reduce grid load.  You have to be sucking a LOT of power to do this, but in exchange for the power company being able to turn you off with little to no warning, you can get a much reduced power rate.  Bitcoin mining being interrupted doesn't affect it other than the loss of mining capacity during those times, so allowing it to be interrupted and restarted for a reduced power rate year round is likely to be a net win as well.

When you have Hydro-electric power you don't deal with these interruptions. 
577  Bitcoin / Mining speculation / Re: The ceiling of difficulty with 28nm technology on: August 20, 2013, 06:48:19 PM
If you are looking for the cheapest electricity costs you should keep in mind that this is just a percentage of the total cost per kwh. The other costs involved are often much bigger than the electricity itself.

If we look at the Belgian electricity cost, at first glance you could think it is cheap, around 5 to 8 € cents. If you dig a little deeper you will find out this is only about 35% of the total electricity bill. You have also distribution costs, transportation costs, green energy costs, taxes, fixed costs etc This will make the total cost of 1 kwh about 18 to 22 € cent, a world of difference.

So is the 2-3 cents the total cost of 1 kwh?

Power interruptions...big difference between (milli)seconds and hours... Can you explain a little more?

I only pay .02 Kwh.  It means everything.   Especially when we forget to add cooling required to hash at the optimal level for your hardware.
578  Economy / Lending / [ANN] Distressed Mining Assets Fund on: August 20, 2013, 06:44:40 PM
Basics:

Create a fund to wait for existing mining hardware to become distress because of the large difficulty increase.   I currently own and operate a hosting facility that can handle this hardware.   Hardware would be hosted with my company and the fund would operate on the 2/20% principle.    Management fee would be 2% of principle and 20% of the profits.  The rest would be evenly distributed to participates based on their level of involvement.   


If you are interested in funding this project, PM me.   
579  Bitcoin / Hardware / Re: A word on profitability on: August 20, 2013, 06:34:11 PM
I believe the mining community is about to go through a real tough spot and once we get through it, we will overall have less miners but more hash-rate.   Basically we are going to see operating costs for miners get to a point where we will see miners that have higher operating costs get bumped out and new hardware will absorb their hashing.  This will produce more concentration of mining and people who have scale and low operating costs will win "at" the expense of others.  Zero-sum game.

Bitcoin mining went through this sort of thing a while back when the price crashed down to $2.50. Certain GPU rigs were now costing more in electricity to run than they were generating in Bitcoin. However, the price rose again and there was new life (very long life as it turned out) for those GPU mining rigs.

Another rise in the exchange rate could buy more time for obsolete hardware.

Maybe, more likely is that new hardware will absorb it so quick you will not even notice. 

Ask yourself this, what would drive a major exchange rate increase at this point?  That is an important question.  If we have no driver that is real then it would just be speculation and that never ends well.
580  Bitcoin / Hardware / Re: HASFAST warranty just TEN DAYS after delivery? on: August 20, 2013, 06:31:42 PM
To be fair with the company not knowing how you are going to treat the hardware and the amount of over-clocking you will do, 10 days seems sufficient to make sure its mining.   With difficulty rising, people will be doing whatever that can to get the most out of their hardware and likely it would be things that would typically void a standard warranty.   

Maybe you should ask them for a official response to your question so we can read their perspective on why they decided to do this.


Question:  How long do you think they should warranty the devices?
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 [29] 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 ... 108 »
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!