What to do with BFL order? Upgrade to hosted or keep jalapenos?

[evlew]

I have an order for 2 7GH/s Jally's from 4/21 which should be shipping any day now. 

My concern is that I think it's still worthwhile to mine with them since I consider my initial investment a sunk cost. However speculation shows that upgrading to hosted while be around twice as profitable if BFL gets it up and running within 3 months. 

The question is, what would you do?  Keep Jally's and mine in hope for a future trade in value?  Or upgrade and wait again?  We all know BFL timelines are shifty....

(Just to clarify... the question is not "What should I have done?")

Thank you for any helpful insight you're willing to share with me.

[1714026973]

1714026973

[1714026973]

1714026973

[1714026973]

1714026973

[Cluster2k]

BFL's hosting plan is still a great unknown.  There are going to be maintenance costs eating into your profit and the contracts are for 12 months.  Presumably you need to pay again after that.  Then there's the risk of BFL launching their hosting product late.  It's already November and in theory BFL should be shipping Monarchs this month or they'll miss their promised launch date.

The one factor working to your advantage in keeping the Jalapeno orders is you know you will be getting them soon.  ROI may be a very long way away (the assumption is difficulty will not keep rising 30% every 10 days into infinity, it has to stop somewhere).

You can buy 'cloud' mining already at places like cex.io.  It is double the price of BFL's offering, but cex.io works immediately.

[bitboyben]

Do half and half. Diversify! That's what I did.

[Perseus353]

Upgrade it all to hosting.

You can get 5x the hashing power for the same money, and none of the electricity or hassle.

I was very happy when I saw the option to transfer my order of a 25 GH Little single to 125 GH hosting.

Yes, difficulty will be up more by january/february, but the growth will be linear from now on -- and it's not mathematically possible to be going up 40% exponentially forever. 

[fizzmine]

Yes, difficulty will be up more by january/february, but the growth will be linear from now on -- and it's not mathematically possible to be going up 40% exponentially forever. 

2 mistakes.  1 you assume BFL will actually deliver by january/februrary, that's a brave assumption.

Hash rate will continue to rise exponentially, but not as fast as in recent months.  The growth we're seeing is ASIC chips starting to catch up with mainstream chip technology.  Mainstream CPUs have been on the 22nm process since Q2 2012, Bitcoin ASICS aren't even there yet.

Intel is projecting 5nm technology in 2019, Bitcoin ASIC could well be caught up by then.

So we will still see exponential growth, but that growth curve will slow the closer the ASICs get to the limits of current microprocessor technology.

I think we'll continue to see large jumps here and there over the next year, but the growth curve will become a bit more predictable as the hardware manufacturers establish themselves and begin consistent product R&D and delivery cycles.

[Perseus353]

2 mistakes.  1 you assume BFL will actually deliver by january/februrary, that's a brave assumption.

Yes, true.

Hash rate will continue to rise exponentially, but not as fast as in recent months.  The growth we're seeing is ASIC chips starting to catch up with mainstream chip technology.  Mainstream CPUs have been on the 22nm process since Q2 2012, Bitcoin ASICS aren't even there yet.

Intel is projecting 5nm technology in 2019, Bitcoin ASIC could well be caught up by then.

So we will still see exponential growth, but that growth curve will slow the closer the ASICs get to the limits of current microprocessor technology.

I think we'll continue to see large jumps here and there over the next year, but the growth curve will become a bit more predictable as the hardware manufacturers establish themselves and begin consistent product R&D and delivery cycles.

It's true -- With KNC and gambling I made ROI on all the mining investments, so I'm not relying on BFL -- it's an extra if any of it does anything.

Anyone know how much PH/s BFL is intending to add to the network with the Monarchs? What about Cointerra?

5nm ASICS will be quite something....

[ScaryHash]

5nm ASICS aren't going to happen before Intel or IBM actually make the technology mature.

As for BFL, why would you keep pi$$ing away money at them, when they've consistently not delivered on time, lied to their customers, and showed to be completely incompetent in terms of engineering and delivering a product on time?

I consider my Jalepenos that have not shipped yet sunk costs also, but I'm not going to be continually sinking money into something that's already sunk.

I figure when they get here, I'll just add their hash rate to the ScaryCluster, and be done with it.