Over the weekend, there was about an hour of power outtage - when it returned all 7 antminers returned to hashing normally but the bitfury did not due to another TOTALLY corrupt SD card (wont format and image - throws 'semaphore timeout 121' errors like the last sd card that became bricked.

off to buy a pair of class 8 cards now see if they handle better.

^this. also, TD knows that this is a new, sensitive category. They do not want 'the internet trolls' to come through and leave a ton of shitty reviews on a product they do not purchase

~1190GH/s for the stack, and more hidden behind it

^sha-256 besides bitcoin are all basically worthless and impractical to mine. Personally, I am planning to mine with my antminers until the bitmain price is in the 1BTC range, then trying to sell them around 1.3-1.5BTC since that seems to be the appropriate markup

here is my rig - 2 antminers and required PSU(s) per crate, plus another antminer on the table behind it.

I am still looking for a suitable location without having to jump right into an empty 1000sqft space right away - Ive had some pretty mixed results on trying to get a good price and the power availability required without  having to go outside the GTA.

Also, I have been trying to wrap my head around starting this as a business, and it looks like there are two obvious options for a canadian small business: sole-proprietorship or as a few shared owners. From my stand-point, sole-proprietorship seems like the straight-forward option with all funds travelling through MY company's account - but If there are some serious individuals who would like to meet in person and share some of the effort and costs of starting this up I am totally open to the idea.

The tax law regarding bitcoin, espescially equipment depreciation, is quite open-ended. My next step is to speak with an accountant or two and see what sort of basic business plan can be drafted up to keep track of equipment costs/profits and make sure capital gains are properly taxed and any small-business benefits are sought out (such as tax-deductions for mining equipment and salaries).

I am still quite certain that there will have to be two seperate approaches made:
1) for investors to back equipment purchases to be hosted on their behalf, and
2) for miners looking for a place to operate existing hardware.

for right now, I am in talks with a small company that might be able to share some room in its back warehouse area which is mostly assigned to storage space at the moment. The location has a 600V 3-phase mains so adding power should not be an issue if there is enough space. I promise to have an update on this option eatrly next week, but it would only act as a small/medium hosting farm, possibly opened up to the 'investors' category. The long-term goal continues to be a larger space that can house tens to hundreds of terrahashes, but would require a large initial investment (hosted antminer group buy?) or a short-term small farm to test the waters for interest and as a business.

I am sure I waited long enough. One occasion i waited about 3 minutes first

on at least 3 occasions 'sudo shutdown -h now' has messed up the SD card image. in contrast when i first had the system i must have safely turned off and on the PSU to do tweaks and pencil mods a dozen times before the first issue with an SD card. At this point i just avoid touching it for fear of it losing my settings.

how could i copy the settings (or better yet, the full image) over the network or onto a usb stick to make backup SD cards?

hahah - maybe meant there were other 2011 or 2012 coins in the safe as well?

is that a *gasp* - REDEEMED - casacius coin!?

sounds like a powewr supply issue. some supplies dont handle the rated power and if thats the case can cause themselves or the machine to switch off or even power cycle

this. It seems like the new fans have a different starting voltage then before, so wont trrigger until the PWM feeds a certain amount of power to it.

I think bitmain is right on the nose with thier pricing 90% of the time. the current 1.45BTC batch and the 2.65BTC batch they had 2 weeks ago both fall a little norht of what i consider 'reasonable'. 0.0065BTC/GH seems like a fair number to me for an antminer (2w/GH) and maybe 0.0075BTC/GH for a bitfury due to the slightly better efficiency.

however, the difference in user experience is vast. the antminers dont destroy SD cards. My bitfury has been stable for almost 2 weeks, but I am terrified to change it over to a less-noisy PSU for fear that the SD card will need to be formatted and imaged all over again just to get it running, not to mention re-doing all the autotune

we are seeing roughly 20% jump every 10 days. If ACTM has wafers by late april it will probably take until the END of may to get them packages, mounted, and through certification/testing if there are no holdups. That is roughly 120 days from now = 1.20^12=8.916 times larger than the current network. Depending who you ask, 20% per jump may be a conservative number, hence my rounding to 10x the current network size.

I just went through all my miners and calculated the error rates - most are between 0.15%-0.9% and running in the 375/387.5/400 range. I tuned one down that was making 2.6% errors and tuned two up that were less than 0.1% errors - presumably that should afford me an extra 10GH or so in total

My immediate impression is that your 12V DC power is far too low. Equipment will not handle well outside of a 11.7-12.4V range and it is almost certainly your PSU being a bit too weak for the job, or running long 18AWG cables that are losing voltage across the distance (partially as the cables getting warm).

I *imagine* that the all 3 capacitors should act the same way when tested - its possible that one might be poor quality and shorting/draining when tested with the meter.

2ND THOUGHT: I am not sure how the power is given from the 12V in to the 4 'segments', but if it is a chain across the the 3 capacitors in a rightward direction the capacitor could be broken and the multimeter is causing the connection to complete, but designed not to carry the required amperage through its probes -> hence voltage drops without a valid circuit. I am not an electrical engineer though, so a 2nd opinion would be a good idea before you take the soldering iron to it. If it is the capacitor, it would be a very easy fix if your friend has the tools for fire-point soldering and some experience with phones. (I personally have only a larger iron so I would not attempt a smaller component like that without buying a new tool or two)

By the way, which section is the one showing all x's? The second one I imagine?

I am given the impression that they may be a bit late joining the party, especially if there are any unforsen delays or problems in the chip. Hashfast already demonstrated how long it can take from the photo of a dorky CEO holding a sealed box of wafers to actually putting equipment in the hands of miners. (Without checking, I think it was close to 5-6 weeks). By that time there will be a lot more 28nm options available (probably including bitfury if they are working on something in the background) and big promises on a 40nm by asicminer (ie: very cost-effective) - and the cost of hardware could begin to drop faster as buyers simply can't operate any larger a cluster at home

okay - I have access to a box of ~20 of them at $5 each but seeing as how 2 failed on me (one mightve been bad from the start and i didnt test right before modding, the other after ~24 hours of operation) I am slightly skeptical. I have a few more to wire up and try with but they seem like great candidates for solid power, with enough for 2 antminers at 375MHz it seems

so ~320TH deploying late may on 55nm -> good news but hardly amazing. If difficulty climbs 75% a month till then, difficulty will be slightly less than 10x what it is now.

For reference I am running 1.63TH of equipment NOW -> rough equivalent of 16TH by the time actm can deploy 20x that (if they meet schedule)

how do you load-balance server PSUs, specifically ones like the DPS-800GB? Can I simply tie all the positive leads to one terminal block and all the negatives to another and get n*800W (n being the number of parallel supplies) of available power, or is there another step? Also, I imagine you couldn't balance a 600W and an 800W together because the 600w will fry, or will it work for ~1400W output?

some power supplies dont provide the full 12V at over 70% rated capacity for prolonged time. I had a 670W server PSU that was switching off and on for 5/10 min cylces at 375MHz. Its fine to power only 1 blade with the other supported by an ATX with 750W on 12V that is also running another 387.5MHz. (an ULTRA X3)