It seems to me like there should be some code in the bitcoin network itself with logic than can have a rough idea of the number/size of outstanding transactions and then require blocks to contain a certain number/size/percentage-of-total-block-size of outstanding tx.  I agree that it's a fundamental issue.

High price, efficient gear coming out, and cooler temps for a majority of the mining population equate to high difficulties for the whole winter I think.  This thing is going to keep climbing 'til the halving, because of the factors mentioned and probably because a lot of people are going to be in a push for the 25btc block reward while it exists.

Do either of those companies have a thread that is months long with extensive photos documenting their facilities and operations?

I like my coin to be somewhere I can trust.  Yes, I know the "t" word is dangerous in this environment, but HaoBTC has to be one of the most forthcoming companies out there that offer an interest earning wallet.

A chaos talisman, I love it, lol!  I though the same thing when I saw that pict, it's like a little ball of wires that somehow mines.  GOT WIRES?  


I dig the "thing" to power the 3 S7s, but shouldn't you cover up that copper bus bar to prevent accidental electrocution?  Looks terrifying!  I hope you don't have a cat...

You should try the U3 support thread, probably get more advice there: https://bitcointalk.org/index.php?topic=827356.0

But general consensus is the U3 is a terrible miner and you may or may not be able to get it to work, everyone has different results.  The best advice for the U3 is to by an R1 from ASICpuppy with custom firmware and let it control the U3.

I can't believe the amount of hardware coming online, it's amazing.  I expected one large-ish jump but we've had several now and there's no immediate sign of slowing down.

Early spring will be very interesting to see if we are at the 100G diff threshold and on the cusp of the halving, I can't wait to watch that roller coaster.

I like the target of around 500W, it's definitely usable in most home environments at that power consumption.  3-400W would be more something I would like to see given the electricity rates I'm faced with, but being able to crank it up/down gives a lot of flexibility and is gives us end users something to tinker with.

The recent push with hardware is efficiency - efficiency, efficiency, efficiency; that's the big headline with any new chip or miner that hits the scene.  But look at what efficient software can do, too.  It maximized the effectiveness of the hardware.  Imagine how much more efficient the network would be if every pool took more time to pay this type of attention to their code.

Could the hub be running out of power, like it can't supply enough to the last stick so the stick generate HW errors? You would expect this if the voltage wasn't turn up enough on the stick, but since it's a different stick dependent on how many other sticks are ahead of it then it might be the hub.

Default is 225, the compacs will run at that freq if you do not include any command, but the frequency setting shouldn't have any effect on bfg reporting your compacs as DEAD.  You may have other issues.

Your freq should be set based on the available amperage to your stick.  I think 250 is right around 0.9A, so good for a usb3 port.

Would you really buy a U4?  I wouldn't, not after what a mess the U3 was/has been.  It's only now that I have the R1 controlling it that I have started to enjoy using it, I don't think I would trust another pod usb miner from Bitmain.

Now if they develop a U4/U5 with a built in controller a la the R1 I may be interested, but a stand alone U3 update?  No thank you.

Today I was watching cgminer and saw two blocks get solved with less that a minute between them, then one more a couple minutes later.   Crazy how fast they come sometimes.

My 9.9% is looking like a contender so far...

You could buy a gekko compac stick for $25 and get like 10x the hash vs. that little usb stick you have.  Still effectively 0 chance at solving a block, but a fun thing to mess around with.

It's 100% chance/luck, there is no correlation.

More sticks = more hash = more chances at a higher share.

I unplug/re-plugged the U3 and it came back to life, no clue why it stopped but it all appears ok now.



Actually better than ok, my U3 runs about 5GH/s faster being controlled by the R1 with the same freq/volt settings I was using previous.  Then factor in the 5GH/s I'm getting from the R1 itself and that's like a "free" 10GH/s upgrade to the U3.  The errors on this U3 are always around 5%, but for a long time last night they were just about 1% which I had never seen them so low before, so I'll credit the R1 with that, too.  Bitmain should have marketed the R1 as a U3 controller first, they may have sold one to every U3 owner on the planet!  I don't think I'll make use of its other functions, but as a U3 controller I'm really digging it.

One more question/suggestion for addition to your custom firmware (besides opening port 4029 as I posted above) - why isn't 250 included in the frequency drop down?  That's the max freq supported by the U3 and it seems to me that it should be a valid selection.  Also, 225 is the default, not 206.25.  There's really no reason to include the frequencies below 225, no one underclocks these and there's no reason for them.



225, 237.5, 243.75, 250 should be the frequency choices.

Thanks!

That's a sure fire way to avoid an orphan, solve the very next block and included the previous solve in it!  Well done, kano.is!

Bitmain uses a pool and look how HUGE they are.  Pools reduce variance, it almost always makes more financial sense to mine in a pool.

Try powering it off for 5 minutes and then firing it up again.  Sometimes I get an X and this always has cleared it up.

Thanks for the clarification, Eric.

Zocadas, you posted incorrect information.  Eric answered this exact question 7 posts earlier.


You only "lose interest" on the high yield account if you withdraw early, the other account still accrues interest hourly, it's just how it is reported on the website that has changed.

What are you hoping to accomplish with this?  If it's additional cooling beyond the stock heat sink then I think having the chips back to back is not a good idea.  Stagger them each side so the chips aren't right against each other.

I know how to do it with cgminer, you run two instances of cg and in each batch file you use --usb :1 which means that instance of cgminer will only use 1 usb device and not look for others, then you configure each instance to point 1 stick to the pool you want.

I don't see a similar command for bfg, so you probably have to do load balancing and send 1/2 the hash to one pool and 1/2 to another.