He has incorporated in Montana.
Cognitive LLC
Cognitive Mining LLC

Sam, any idea what happened with the divs?

The deposit was made to the Havelock address and their system says it was paid on the 3rd but no divs were distributed. Could be a confirmation thing or a paying at midnight thing.

Havelock's support people roll out of bed at 9am EST (6 hours from now), you might want to give them a shout around then.

Same for everyone I think.

Blockchain says that Garr made the deposit.

My guess is he set the time for the dividend to be released but the deposit transaction didn't have enough confirmations by the time the dividend was scheduled to happen. There was almost an hour between blocks around the same time. I could be wrong though. I don't really know anything about the issuer side of Havelock.

We'll probably have to wait a few hours for someone at Havelock to wake up and sort it out.

Have you sold all your CT machines yet or are you keeping some?

It would be a lot easier to find if you'd stop deleting your posts  


It could be argued that he's talking about the current results not necessarily the final results.

Edit: And the 27th was 7 days ago.

IIRC, there were 2 motions.
The first was the one with the start date 20th of March.
Motion was "Liquidate YES/NO" and came back as No

The second was started on the 27th to finish today.
Motion was "Pay Hosting/Liquidate". It came back as not to liquidate.

That would only be wise for in hand hardware. Any waiting around would be very costly.
The data center is a big part of the problem but the exchange rate is the killer.

At $270/kWh/month running 1 unit at 1.6 TH/s costs ~ $540. At today's rate, that is 1.2 BTC
At a difficulty of 6 billion (next 2 weeks) and 7.2 billion (the following 2 weeks) that miner will produce 3.44 BTC.
This leaves us with a profit of 2.24 BTC per unit for the next month.

If the exchange rate goes up or we can find cheaper hosting, Cognitive would be in a better position.
In the slightly longer term, reinvestment is going to be the decider (court cases and pissing contests not withstanding).

If we had deployed 30 TH/s on Feb. 18th (when the first CT machines arrived), over the course of three months we would have paid an average weekly dividend of 1.3 mBTC per share and had over 200 BTC available for reinvestment. Now we only have ~20 TH/s with ~10 TH/s waiting for delivery. Recovering from Cognitive's current position is going to be tricky.

Hash rate at the Canadian center started dropping about three hours ago. We're down to around 13 TH/s now.

Yeah, there really should be more communication.
What happened with the attempt to buy btct.co from burnside? That was a much better exchange. Particularly the motion system.

It's an unregulated exchange though. I don't know what (if any) laws would apply here.
But, yes. It shouldn't have happened. This whole situation is a mess.

That's probably what triggered Havelock's alarm and caused them to freeze trading.
I think they implemented it to catch hacked accounts dumping shares like guybrush01's situation a few weeks ago.

Same here. Last week it was on the line. Some careful work could have turned it around. Now, events are working against us.

Hash rate is 22 TH/s now. Shame it wasn't a month ago.

The average price of the last 10 TerraMiners sold on ebay is $7,900. At the current exchange rate, that's 16.81 BTC per unit.

I'm inclined to agree. The two biggest problems are the exchange rate and Active Mining finally going into production.

The low exchange rate means we can get more BTC for our hardware. We pay more in BTC terms for our hosting ($0.37/kWh is pretty steep too). After hosting, the reserve fund won't be able to buy us a significant amount of new hardware and definitely not enough to stay ahead of the diff growth.

The Active Mining situation will bump difficulty growth to the high side of the 15-20% range. We were looking at growth in the top of the 10-15% range until Friedcat's gen 3 hit the market but Active Mining's ~4 PH/s will bump it up in the mean time.

Unless Garr can magically double our hashrate in a few weeks for no money or the exchange rate rises enough to make our dollar costs negligible, Cog is pretty much dead in the water.

We don't know anything. All we have is your word. Want to share your evidence with the rest of us? Or at least let us know where you are filing your suit against Garr so we can see what the courts have to say on the matter.

In regards to voting, we can use something along the lines of gmaxwell's "proof of solvency" idea to provide a verifiable voting record without revealing anyone's personal information or shareholding. Garrett can build the merkle tree, publish the root hash here and email the fragments to shareholders. Vote totals can be verified via Havelock and voters can verify that their votes were included in it.

No problem. It's a complicated situation.

If they were converted immediately (they weren't) they would have been worth that much in hardware. The share price is usually worth more than just the price of the hardware though. COG.F and COG.F2 were essentially a bet that the share price would be greater than 0.25 BTC when they converted which seemed very plausible at the time. There are a few factors that run the price up.

That hardware (if it was running) produces bitcoins, so the dividend yield increases the price of a share.
Acting as a group allows us to make better deals (i.e. the $500 discount per unit) than if we were buying alone, this increases the share price.
Because we have many machines running rather than say just 1 each, a single machine's failure is not catastrophic & the risk is spread.
We have an economy of scale when it comes to other things like hosting, power and management which all add to the value.

As I said in a previous post, historical data for BTC mining securities suggest that people typically value an asset like this at between 3 and 10 times its annual yield (an "interest rate" of 10-30% in BTC terms).
i.e. if the dividend changes, the share price will change to ensure the following is true: 0.3 > ((weekly div)*52/(share price)) < 0.1
This has generally been the case with some exceptions resulting in spikes up and down depending on the disaster of the week.

At the current difficulty, 20 TH/s (if running) will produce ~14 BTC per week which (after reinvestment) yields 0.5 mBTC per week. If there was faith in Cognitive going forward, that would traditionally have supported a share price of between 0.078 and 0.26 BTC per share. We also have another 10 TH/s to be delivered.

Where it went wrong was Garrett waiting too long for the Washington data center. The second Cointerra started delivering, we should have been deploying. There should have been an intermediary datacenter in place beforehand. That's what is being done now. If we decide to keep this show going, we'll have to take measures to ensure something like that doesn't happen again. As I said elsewhere, I expect the difficulty growth rate to slow, which means our current hardware will be running (and paying off) for quite a while. If we can manage even one more good deal and deploy it quickly, Cognitive will be set for a long time.

But all that depends on whether shareholders decide to liquidate now for ~0.02 BTC per share or not. If cog keeps going, it will also depend on us figuring out a better way to manage it. YoYa suggested setting up a board consisting of Finance, Mining and PR which seems like a good idea to me.

I'd like to see those numbers too.



The best I can describe it is as follows (bear in mind, I'm working from memory here so the numbers might be a little off).

In dollar terms:
We bought 14 units for ~$78,000 which is $5,500 each (discounted from $6,000).
We will receive a total of ~22 units currently valued at ~$8,000 each for a total of ~$176,000
This is a 225% profit in dollar terms.

In bitcoin terms:
We bought 14 units for ~650 BTC which is 46 BTC each. [ 1 BTC = $120 ]
We will receive a total of ~22 units currently valued at ~15 BTC ($8,000) each for a total of 330 BTC. [ 1 BTC = $520 ]
This is a ~50% loss in BTC terms purely from the increase in the $/BTC rate.

If COG.F/2 shares were instantly converted to ordinary shares at the time of purchase (i.e. when we made the deal for the hardware), the value of the hardware per share would have been 650/14420 = 0.045 BTC per share. Now it is 330/14420 = 0.023 BTC per share. Which is a roughly 50% drop in terms of BTC.

The profit in bitcoin terms here is clearly not in selling hardware but in mining with it if we have it in hand. You can see those calculations in previous posts.

The problem is that the CT hardware was priced in dollars. At the time 1 BTC was worth ~$120. IIRC, ~650 BTC were needed to raise the required dollar amount. If BTC was still worth $120 and we sold 19 units at $8,000 each, that's 1,266 BTC. Now BTC is worth 4+ times as much. If we sell the hardware for dollars now, we get ~300 BTC.

If BTC keeps rising, it's better to keep mining. The COG.F holders effectively locked in a share price of 0.25 BTC (5 BTC for 20 shares). The dividends from current hardware (assuming all 30 TH/s gets running) could support a share price of between 0.12 and 0.38 BTC based on 50% dividends alone. If the reinvestment fund is used well, it could be higher.

I expect mining difficulty increases to continue to slow down except for a spike when ASICMiner's gen 3 stuff comes online. This means that our current hardware and whatever we buy next will perform for a long time. I'm sure it is possible to salvage this operation. The question is do we all agree and if so, what plan can we come up with. YoYa's board suggestion is a good one and worth following up on.

You'll have to reply to the email he sent out to vote.

Also, we have movement on Eligius.
New machines seem to be running at 1.6 TH/s. Our 256 second average is 10 TH/s.

There's no buyback in the contract:


Liquidating Cog means selling the hardware with the proceeds being distributed to shareholders.

TerraMiners are going for roughly $8,000 at the moment. 30 TH/s is ~25 units at 1.2 TH/s (19 at 1.6 TH/s). This gives us between $152,000 and $200,000 assuming all units are sold. At current prices, that is between 287 and 377 BTC or between 0.020 and 0.026 BTC per share.

Running 30 TH/s assuming a difficulty increase of 15% every 2 weeks will yield 108 BTC in 6 weeks, 203 BTC in 16 weeks and so on. And we'd still have the hardware although it's value will be greatly depreciated.

Note that the above doesn't include Ebay fees for selling or electricity costs for running the hardware.

It's a tough one to call. If it's run correctly, there is still a viable enterprise here. If it's not run correctly, then its assets are worth about 0.02 BTC per share. The last motion to liquidate seems to be "no" by 3:1 but needs to be updated. The current motion to use reserve funds as operational funds or liquidate cog expires in 7 days. I'm really not sure which way to go on this.



Havelock should clear the order book before trading recommences. There'll be no buying at 0.002 BTC unless shareholders decide to put up an ask at 0.002 BTC.

What numbers am I dumping?

There have been times when the share price went through the roof without the returns to justify it. See the jump from 0.5 to 1 BTC last July.

There have been times when the price dropped like a rock because of panic selling. See stabs dumping 400 odd shares into a shallow orderbook a couple of days ago.

You'll see similar movements for most other btc securities out there. A slight exception being ASICMiner.

However, for most BTC securities *in general*,  0.3 > ((weekly div)*52/(share price)) < 0.1
This means that, generally, people have decided that risking their money for a 10-30% reward was worth it.

When (or if) trading reopens for COG and the first dividends have been paid, people are going to try to value a share. History says that the above rule would be true (in normal circumstances), giving a price of >0.08 BTC. But at the moment, most people will be worried that either Garr will fuck up again or Goat will try to tear the whole thing apart. This will have an impact.