I am using 3 phase 480v service. Single phase to neutral (or supplied ground) is 277Vac (actual measured value is around 272-274V). I have a few grounding rods tied together for the PSU ground. And I do have some 240V available, it is just easier to for me to add new 277Vac breakers, this will also slightly reduce the amps vs 240V.
OK, I have been warned. Does anyone have instructions on how to disable the OVP? (Only time I ever blew a psu doing this is was when the selector switch on the ATX PSU was on 110V)
Thanks!
You are creating a serious and real fire risk running the PSU higher than its input rating. I would also not recommend it, get a transformer and do it right. This is not being over cautious, real fire risk, I'll repeat it. |
|
|
|
LMAO, look at one of my older posts, I knew for a fact this facility would overheat based on what I saw... Little do these guys realize, you need about 150cfm per KW of airflow... That facility @ 1.2mw does not have a 275,000+ cfm worth of airflow through 36'' ducts...
What a joke, the concrete structure makes it a heat trap / chimney. The white smoke test they did took several minutes to clear out... We fully cycle our air from intake to exhaust every 6 seconds... Yeah, so a white smoke test in a properly designed facility @ 100cfm per square foot vacating air allows you to keep your units cool. And the smoke test, would simply flow from intake, through exhaust, and be 100% clear in less than 10 seconds..
These guys are attempting to condition air, and not move enough of it, leaving them with massive hotspots, and expensive PUE, and a horribly inefficient system.. The only thing you have is .03c per KW/h compared to china's .04 / .05 cents per KW/h
Can't wait until we go public in some of our assets so everyone understands why they can't compete... I mean, look at Bitfury, they have a facility in Georgia (the country Georgia) directly under a hydro powerplant, paying less than .03c per kw/h, evaporate only, don't have to worry about compressors, and have about 20mw+ of capacity and expanding...
With the changes in heat sink design and methodology as well, you don't even need to treat the air.. You just have to move a lot of air, even if its 90 degrees across the chips rapidly and your fine... At least the direction bitmani is going with there small chip set.
Good luck if you go in this facility.. Why do you think Spoondoolies took one look at the facility and said no way... They decided to go to a different datacenter in Washington and a footprint @ vern global Iceland instead? Bitcoin miners consist of two groups now, large industrial players, and hobbyist, don't expect to make a dime... Its a hobby YOU pay for, not get rich on... You can't compete with my sub 40-50 dollar per KW cost including overhead staffing and cooling... And the fact that in massive orders, we pay about 50% of what normal people pay for equipment from the manufacturers due to large scale contracts.
If you have free power, and live in a cold place, YES! you can make money, these guys can't even make money if they are by themselves building this facility with compressors and attempting to treat air. I'm sure the capital for the contractors they used too was excessive as well, that looks like a very expensive facility that can't even run at 60% of its promised capacity during the hot months... You guys are lucky its cooling down, you'll have 6 months to get your shit together.
We can compete with you and everything is properly cooled and is running on quality infrastructure. You make some great points. We choose to grow slower but build better and still offer the best price. So you're telling me, that you can compete, with my facility, if I were for a significant profit, offer $50.00 per Kw per month hosting once we finish our 5mw expansion for possibly hosting other large clients? FYI, we used remote power switching PDU's, tied into a miner monitoring software that automatically reboots the miners if there is a problem. If the unit doesn't come back up after a failed reboot, then an alert is setup to the NOC and to our guys at the DC for them to work on the unit... We run @ about 99.95% uptime from all our units. Even miners that freeze, or have blades go down every 12 hours still run at about 95-98% efficiency because of our ability to monitor and remote reboot the units... This allows me to reduce the amount of labor costs. Oh also, in an emergency if we loose an evaporate cooler, the system can automatically load shed (turn off the units) before they begin to overheat and cause permanent damage to the ASIC's, an event you had earlier in April... Our goal is to offer mining contracts to people, not host miners, direct, as we can purchase units at 50% of cost wholesale, and actually pass on real profits to individual customers, while still making margins on the hosting and the unit / resell, but offer power rates equivalent to $50 per kw instead of $60, and once cap X is paid off in a couple of years, we will have operation costs in the $30 dollar range with maintenance and personal. What mining company do you have an agreement to get at 50 percent of wholesale? That is a heck of a goal. I would sure be interested if you can get units for close to that.... but I'm not sure I see 50 percent happening. This is also first time I have seen hosting companies argue.... very interesting  No argument here, just stating a fact. Who wants a provider in a race to the bottom? At a point something is sacrificed. No free lunch. |
|
|
|
Bank on $500,000 - 700,000 per Petahash and you won't be too far out.
That assumes you can find a company willing to sell you the mining equipment at a reasonable cost and don't forget about physical and internet security systems.
This is a decent estimate. |
|
|
|
Spot on with the article, thanks for taking the time, good material, good refresher. Definitely points to some solid points as well as good food for thought. It's fun to look at miners that have been around for about a year; yet their inefficiencies make them obsolete.
Thank you. Any suggestions for another article you would like me to cover? I am brainstorming ideas for the next article. I want to get down to writing. |
|
|
|
LMAO, look at one of my older posts, I knew for a fact this facility would overheat based on what I saw... Little do these guys realize, you need about 150cfm per KW of airflow... That facility @ 1.2mw does not have a 275,000+ cfm worth of airflow through 36'' ducts...
What a joke, the concrete structure makes it a heat trap / chimney. The white smoke test they did took several minutes to clear out... We fully cycle our air from intake to exhaust every 6 seconds... Yeah, so a white smoke test in a properly designed facility @ 100cfm per square foot vacating air allows you to keep your units cool. And the smoke test, would simply flow from intake, through exhaust, and be 100% clear in less than 10 seconds..
These guys are attempting to condition air, and not move enough of it, leaving them with massive hotspots, and expensive PUE, and a horribly inefficient system.. The only thing you have is .03c per KW/h compared to china's .04 / .05 cents per KW/h
Can't wait until we go public in some of our assets so everyone understands why they can't compete... I mean, look at Bitfury, they have a facility in Georgia (the country Georgia) directly under a hydro powerplant, paying less than .03c per kw/h, evaporate only, don't have to worry about compressors, and have about 20mw+ of capacity and expanding...
With the changes in heat sink design and methodology as well, you don't even need to treat the air.. You just have to move a lot of air, even if its 90 degrees across the chips rapidly and your fine... At least the direction bitmani is going with there small chip set.
Good luck if you go in this facility.. Why do you think Spoondoolies took one look at the facility and said no way... They decided to go to a different datacenter in Washington and a footprint @ vern global Iceland instead? Bitcoin miners consist of two groups now, large industrial players, and hobbyist, don't expect to make a dime... Its a hobby YOU pay for, not get rich on... You can't compete with my sub 40-50 dollar per KW cost including overhead staffing and cooling... And the fact that in massive orders, we pay about 50% of what normal people pay for equipment from the manufacturers due to large scale contracts.
If you have free power, and live in a cold place, YES! you can make money, these guys can't even make money if they are by themselves building this facility with compressors and attempting to treat air. I'm sure the capital for the contractors they used too was excessive as well, that looks like a very expensive facility that can't even run at 60% of its promised capacity during the hot months... You guys are lucky its cooling down, you'll have 6 months to get your shit together.
We can compete with you and everything is properly cooled and is running on quality infrastructure. You make some great points. We choose to grow slower but build better and still offer the best price. |
|
|
|
|
Save what I can for the future.
|
|
|
|
|
No, perverse incentives attract perverse outcomes.
|
|
|
|
I'll just hire 10 asians to blow on the miners as they go.
10/10 would asian labour again.
What a racist I believe he was joking. |
|
|
|
Hey everyone,
I'm a little confused on this subject. Bitcoins are limited and about half or a little bit more have been found already. I'm confused on the part when all bitcoins have been found what happens next? Will there still be a way to profit bitcoins by mining? Or is there something I'm not getting right? Let me know, Thanks guys.
you will most likely be dead. why worry? the reason i wrote this is every 2-3 years blocks 1/2 so 50 two 1/2 years 25 two 1/2 years 12.5 two 1/2 years so five years in 6.75 two 1/2 years 3.375 two 1/2 years ten years in i think we last into 2105 or more i will not be around and most others will be gone. why worry? Is this a joke ? https://en.bitcoin.it/wiki/Controlled_supplyThis is not a joke, but a valuable feature of Bitcoin. |
|
|
|
Duplicate, also needs date formatting to fit into the news section.
Thanks...
Agreed, please read the posting rules and look at how they are formatted before posting. Thanks. |
|
|
|
I have no idea what drives the price of Bitcoin. Yes, I hear "Supply and demand", but I have zero idea, nor have I heard a good explanation of what drives Bitcoin demand. While I am general believer that "difficulty follows price", that doesn't really help much.
As my knowledge, Bitcoin price is based on Mining Equipment Investment and Profitability. Not at all. Then price would not have reached 1000+ in 2013 and fall below 200 in 2014. Like any other free market asset, Bitcoin price is dependent on complex trading functions like future, shorting etc. The price for the most part is based on the cost to generate a coin. That has to be in balance with the trading markets or you will get into a situation where the market price is lower than the cost to create a Bitcoin. If that happened for a prolong period of time you would see the difficulty decrease until it was back in balance. In the real world this has happened with the paper gold price (Comex) versus the physical price. When the miners have the paper price dip below their cost to mine the same amount of gold they stop or severely slow down production until the price rises so they can cover their costs. This assumes they are un-hedged. Same thing is happening in the U.S. oil market with the frackers. Many of their hedges at higher prices ($60-$80 dollars per barrel) are expiring so now they are faced with producing at $40 per barrel which is unprofitable. This can be proven by looking at the well and rig count in U.S. oil fields. They are declining and as more hedges come due and are not renewed at that higher price, more capacity will be brought off the market. |
|
|
|
i mean you do the mining get the hardware calculate the ROI to be one year and after one year you get the money back and a deprecated miner to sell like 1 dollar per 10Ghs power and a 10TH miner sells like 1000 dollars into 2016 so you get 3 dollars per day into what you invested and if watts not too expensive... i mean its a surveillance every day for just 3 dollars of a 4000 dollars investment its feeble.
Pretty broken English but I think I understand the nature of your question Tony. Yes your ROI horizon needs to be at least 6 months to 1 year. And yes, you are speculating on the future price of Bitcoin for the most part or you are selling most your coins to covers the expense (hosting) of mining. If you don't believe Bitcoin's price will be higher in the future then you should only be mining based on reasonable short term ROI numbers. The mining herd is being thinned right now and this process really started when the first FPGA and ASIC chips hit the market. |
|
|
|
A minor editorial correction: I think you meant 400 Petahash rather than 40 at the end of the first paragraph.
I am also curious about why "retail sales" would be excluded. I get that Bitmain didn't want to rely on retail sales to cover the costs associated with developing and producing the S7 (as an example). I expect they figured that their internal mining operation would cover it no matter what, and that retail sales probably have enough additional margin that it only enhances their finances. They sell however many hundreds or thousands, and they have transferred almost all the risk to the customer. Seems like a win for them.
Granted they may charge such a huge price that retail sales decline, but it seems that for a while anyway they aren't hurt by retail sales at all.
Thank you, I didn't hit the zero enough times. Fixed. Bitmain will do this for this generation and maybe even their 14nm-16nm offering but nothing more. Too much money would be at stake and the market for it would be much smaller in size but each customer would be larger. That is why you will see only private invite only sales for any large amount of capacity. Maybe we see some crumbs come to the public but I think that will be the extent of it. Mining is consolidating and this is intensifying. We you are looking at a minimum $10 million dollar bill for the next ASIC chip on a small die, you really need to make sure you are covered or there will be hell to pay. I wish I could think that the public will be able to support this development but in the end, the incentive just isn't there for small miners to cough up the cash needed to cover that. Only pre-arranged sales or 100% private mining will do it. |
|
|
|
The the reward halves we will see either a rise in prices to bring it back into marginal profitability or the difficulty will drop because of more miners dropping out that can not mine Bitcoins at that reward level.
Thats the paradox, in my opinion: we think price should rise and be happy about upcoming reward halving, but - see what happened to Litecoin when reward halved: price tanked. Same thing happened to many altcoins when reward halved, it seems to make the coin less interesting so price goes down, not up. Scarcity makes a coin less attractive it seems: maybe a more detailed study can be made regarding this. Altcoins and Bitcoin are completely different in this one single important fact. Somehow, Bitcoin has gained the status the not many items gains, it has human confidence in it. Like real confidence that Bitcoin has value and no matter what will continue to have value. At what price? Anyone's guess, but a price none the less. This means Bitcoin can operate outside of just hard data. At any moment this could cease and what I am saying would be incorrect. Remember halving to Bitcoin means we are getting closer to the end of the fixed supply and in that is where Bitcoin is different than any other currency in existence. As long as no one increases that amount to be issued, that special feature will continue to buoy the currency. We will have to either see one of two things happen when we see the next halving, price just about doubles if it is at current levels or difficulty will lower to get back in balance which would mean a drop by as much as 50%. One or the other will happen, maybe not right away but you will see a trend in one of those directions. |
|
|
|
Very interesting article, congrats! There is also the impact of future reward halvings of BTC mining that has to be factored in, which effect I am unsure of.
So far every coin's reward halving (e.g. Litecoin) has paradoxically been met with a dropping price, which is counterintuitive because of the reduced offer.
I think that may be that the profit of mining is such a small margin for most miners that they are always selling Bitcoins to covering their hosting expenses. That would put constant selling pressure in the form of newly minted Bitcoins always being for sale. The the reward halves we will see either a rise in prices to bring it back into marginal profitability or the difficulty will drop because of more miners dropping out that can not mine Bitcoins at that reward level. I glad you enjoyed the article. I am constantly thinking about things to write about, more to come, working on my next piece already. |
|
|
|
|
Get some HVAC rated filters and you need to build a manifold that can house enough for your desired air flow and make sure you are "pulling" across it, not "pushing" through it.
|
|
|
|
Hmm I think the link is dead but it might could be my bad timing. Please check and update. In regards to your comments about immersion being the most expensive, did you mean submersion or 2 phase immersion cooling techniques from Allied Control? Also given Bitfury's recent acquisition of Allied Control, I'm eager to see how this concept is adapted to such an extreme scale to meet Bitfury's super datacenter requirements. I imagine it will put the proof of concept design built with ASICMiner a few years ago really to shame.
Fixed. Immersion Cooling - Most expensive option overall at this point because there is not scale in the production of immersion cooling tanks and supplies. I was referring to Allied Control's setup being that mineral oil is not fun to work with. Allied Control uses a liquid that is not oil based and it much easier to work with the clean up. ASICMiner was using Gen 1 of Allied Control. They are the leader in this new area. Over time the cost will come down. We are looking to invest in it within the next 2-4 years. |
|
|
|
|
Glad to see new hosting companies stepping up and doing more than asked to give miners confidence in them. We need more good actors. Cheers.
|
|
|
|
Quick question.
If I use an ethernet switch in my room and attach an ethernet cord to the switch, then attach a cord to my desktop and an antminer s5, will my internet connection with my desktop slow down? I don't think it matters if it is a gigabyte switch or not seeing how only 100mbs per second reaches the switch. Thank you.
You should see little difference. The only thing I could say is if its a cheap combo router from your ISP, it could have overheating problems that may freeze up you wifi is its built in. In that situation I would disable the wifi, get a standalone wireless modem and run just the switch on the router. The best would be to not use the router at all for switching and get a 8 or 16 port switch and use that for your network and the router just for serving up your connection. |
|
|
|
|
Show Posts
