Two GTX GPUs on a 650 is pushing it, even before adding in all other components (CPU, RAM, HDD, case fans, cd/dvd drive, etc). You always want to leave a bit of room in your PSU anyway...if you use 600W, it's better to get a 700-750W PSU than a 650W.

I've used this Power Supply Calculator every once in a while, though unfortunately they haven't updated their lists to include 280/290s yet. It still provides an idea of what you'll need overall though.

I apologize if this has been asked already, but has anyone noticed any significant variance on either of the boards (from the status page)?

The first board, CTA0, is varying pretty wildly at higher power steps (~150GH/s up through 800GH/s) with quite a few errors (786 out of 378,000...not so bad, but way higher than CTA1). The only issue I've found so far is that out of the 8 cores on that board, 1 is reporting "0" voltage (all others are stable around ~620-630). Normally I didn't think that would pose such a wide swing in hash rate.

Anyone else with a [possibly] dead core?

I've got quite an update for this evening.

1) A coworker of mine has been interested in Bitcoin ever since I introduced them to it last September. Since I needed to move the old BFL Singles out of our datacenter, he offered to purchase them from us for BTC1.7, to which I've agreed. Given how the network is beginning to ramp up with the new hardware being shipped, I figured this was a fair price, and my coworker didn't mind. Payment will be added directly to our 'hardware reinvestment' address.

2) Our January CoinTerra order has arrived! I spent quite a bit of time today rearranging our portion of the rack to accommodate the size. Here's the unboxing album.
For those of you wondering, you can change the power stepping on the rig from 1 (the minimum power usage) to 9 (the maximum). At the minimum, the rig uses a bit over 1000W and runs at 800-850GH/s. I've been slowly stepping up the power to make sure I don't cause any unneeded disturbances / fluctuations. As of right now, I've set the stepping to '5', and we're running stable at 1.1-1.2TH/s.

You can see us on the BitMinter Live Stats page for once, though I'm positive our tenure there will be short.

Our combined rate is currently 1.8-1.9TH/s.

EDIT: I've moved the power settings up to 7-8, but had no discernible impact on our hashing speed other than just using more energy. Apparently this is a known issue, at least according to the CoinTerra thread. For now, I've set it back to '5' (1.1TH/s) until I hear back from CoinTerra on a ticket I opened with them.

To me, the question "Who will be left after all the dust settles?" implies that the remaining companies will have the reputation, knowledge, dedication, and financial capacity to last seemingly indefinitely (under the best circumstances of course).

With that in mind, I'd say my primary pick is ASICMiner, solely based on how far they've come already...and assuming they continue to develop chips that outpace competitors.
Choosing a second company is difficult though. I'm not sure any of the other options truly fit the bill, but KnC, CoinTerra, Bitmain, and BitFury are close.

And then there's BFL, a Bitcoin-parasite who will be around so long as their lawyers don't leave any loopholes for them to be sued by, all while slowly draining the wallets of any newbie who can't use a 'search' function. 

In other words, mining is *not* a thing of the past, which is what I was trying to get you to point out.

But yes, whenever or wherever there's a chance to make money (even with slim margins), large organizations will slowly migrate over to it, consuming or crushing the small players in the process.

Oh? Replaced by what?

Lol looks like BFL has some direct competition this time around with their PCI cards  

https://cointerra.com/product/gsxi/

Looks like they'll be releasing about the same time as BFL too


CoinTerra's response to BFL's nonsense:

You're talking about fiat when you should be talking about Bitcoins. Fiat price doesn't matter because of the amount of time it takes to get cash into Gox to begin with.
It's a very likely possibility that Gox will allow Bitcoin withdrawals to external addresses, the question is just 'when'. Speculating 90%, for example, implies a full-faith mentality that Gox will be up and running shortly.

When all is said and done, one person will have more Bitcoins, and another will have less. It's all in how you value risk.


That's a bit childish, don't you think?

Willing to buy GoxBTC @ 70% (0.7 realBTC -> 1.0 GoxBTC), though I'm happy to negotiate.

I prefer to deal with established members.

Sounds good, and happy to help. I'll play around on the site some more and email you guys directly if I come across any other issues.

Cool beans, I'll have to check it out.

EDIT: I'd like one infinity bitcoins please!

Brief update for everyone:

We've sold one of the Singles for BTC0.55, which I've considered a fair market price.
For those of you wondering, I price hardware based on:
A) What currently available hardware there is out on the market.
B) The efficiency of both the equipment I'm selling and the equipment on the market.
C) What new hardware is coming to market over the next few months.
D) Difficulty projections.

I *Do Not* go by the ridiculous prices I see on eBay.
I aim for a breakeven period of 60-days. After escrow and shipping fees, approximately BTC0.5 will be added to our hardware reinvestment address.

On another note, we received our BitMain AntMiner S1 today, an order I placed just 1 week ago. Everything seems to be up and running quite well so far at ~180GH/s.

Overall, we've seen a net gain of 120GH/s for this month (lost 60GH/s for the sold Single, but gained 180GH/s from the AntMiner). It certainly helps, but not as much as seeing our CoinTerra gear arrive.

I'm a fan of KnC equipment, but the uncertainty behind the Neptune is really what's holding me back from considering a purchase. A "Q2" ship date and unconfirmed specs are my primary causes for hesitation, especially given what I've seen with CoinTerra so far. Hopefully new info comes out soon so I'll be in a better position to make a decision.

That's correct, Havelock current manually confirms all withdrawals from fund accounts. Since there shouldn't be any reason to withdraw coins from BDD, the likelihood of this being approved is fairly low.

And if I remember correctly, that 1% is actually 0.4% (Havelock fees) and 0.6% (his commission) [NAV increase].

EDIT: Ah, I was mistaken. Thanks for the clarification guys.

Fantastic news, thanks for the update Friedcat! You should be receiving a PM from me shortly. 

Other than what they've posted on their site, I haven't heard/seen anything else unfortunately. Hopefully they beginning shipping shortly, though they'll need to get through December orders before our January one.

Another disappointing fact is the power consumption. I'll need to get in touch with my CoLo manager and see how large the circuit is that I'm using... a few BFLs and a KnC is one thing, but 4kW over 2 4U units will undoubtedly raise my monthly rates. Ugh.

Never go into mining assuming you're going to make a return on your investment within a month, 2 months, 6 months, etc.

And, more importantly, never purchase from Butterfly Labs. They may market themselves better / more widely than their competitors, but their products A) under perform, B) costly per GH/s, and C) are never shipped on time (by a factor of 4-6 months or more).

Having just found them, I'd probably sit there in a state of shock for a bit.

After I get my mind right, I'd withdraw 5% or so to fiat to cover RL expenses for a while, another 5% set aside for misc. purchases (think Overstock, Tigerdirect, or anyone else accepting BTC), 25% reinvested into Bitcoin projects I've got on the table, and the remaining 65% set aside in cold storage.

You had it set for 150TH/s, not GH. And it's also assuming a consistent growth of 78%, which is incredibly unlikely.

Some quick 'napkin' math says 25TH/s for BTC1 would require the difficulty to be, what, around 350B or so (network hashing speed at ~3500PH/s).
That's essentially the equivalent of selling 1.75 million Cointerra TerraMiner IVs, worth ~$10 billion.

Impossible? Definitely not, it could certainly happen. But over 12 months? Probably not. Bitcoin has a lot of growing to do before it reach that point.