here's just how screwed ASIC buyers are - READ THIS if you have a preorder

[DutchBrat]

And we all know that the cost of the ASIC devices will plummet as the marginal cost of producing the chips after the initial investment is close to zero (or very low)

Everyone seems to be throwing that around but I don't see it happening.  The 3 manufacturers almost have a monopoly.  BFL will release first in all likelihood and they have a monopoly on "budget" hardware that's a bit slower but more affordable.  bASIC will have a monopoly on larger rigs that are actually built well.  All the pics I've seen suggest BFL uses REALLY sketchy parts like cheap fans and stuff.  And Avalon is releasing their products so unbelievably late, they're not even in the picture.

So that means they can all charge whatever they want.  Here's my theory.  Once a company has sold XX amount of mining hardware up front, jack up the price.  That's because for every ASIC miner you sell, you devalue your product.  The less people mining, the more you can charge because people's ROI calculations will result in a smaller time period.  Let's say it takes $50 to build a Jalapeno.  If the price is $300, less people will buy it but they'll be willing to pay more seeing as how less people are buying them.  Then it's just a cost of manufacturing thing.

You contradict your own post where you say that people buying an SIC now will not make their money back...

So how does increasing the price afterwards make more people buy your product if they can't even make their money back on the earlier bought models....

Apart from Avalaon, bASIC and BFL there's also Tycho coming with an ASIC in Q1 2013 and ASICMINER will be heads on with BFL in rolling out the first ASICs albeit they will use the 1st 15TH wordt of chips to mine for their shareholders and then start selling them.

With an increasing network speed prices will have to come down or no-one will buy

[1715006599]

1715006599

[1715006599]

1715006599

[1715006599]

1715006599

[1715006599]

1715006599

[1715006599]

1715006599

[Sitarow]

And we all know that the cost of the ASIC devices will plummet as the marginal cost of producing the chips after the initial investment is close to zero (or very low)

Everyone seems to be throwing that around but I don't see it happening.  The 3 manufacturers almost have a monopoly.  BFL will release first in all likelihood and they have a monopoly on "budget" hardware that's a bit slower but more affordable.  bASIC will have a monopoly on larger rigs that are actually built well.  All the pics I've seen suggest BFL uses REALLY sketchy parts like cheap fans and stuff.  And Avalon is releasing their products so unbelievably late, they're not even in the picture.

So that means they can all charge whatever they want.  Here's my theory.  Once a company has sold XX amount of mining hardware up front, jack up the price.  That's because for every ASIC miner you sell, you devalue your product.  The less people mining, the more you can charge because people's ROI calculations will result in a smaller time period.  Let's say it takes $50 to build a Jalapeno.  If the price is $300, less people will buy it but they'll be willing to pay more seeing as how less people are buying them.  Then it's just a cost of manufacturing thing.

In 6-12 months those ASIC devices will be replaced with better equipment.
During that time they will not lower their Price and here is why.

Reason 1. Have a product to offer that is better then your competitor makes you money.
Reason 2. During the time of development they will have to fund that research.
Reason 3. If you lower the price of the devices your existing customers bottom line will be effected.

Solution 1. Reveal a 2nd generation device before your competitor.
Solution 2. If it makes business sense offer full value trade-in on  first gen pre-order devices.
Solution 3. If you happen to lower your prices. Offer trade-in options will less credit towards 2nd generation pre-order devices.

IMHO To have a successful business and or project you should adopt a policy that will never burn your bridges with your customers.... Ever...

[Richy_T]

If you go through the order process and never fwd the bitcoins or payment the order number still goes up by 1.

Ohhhh, they should clean that up, lol.

The customer should never see the order# (or at least the row number in the database). They understood this even back before computers were in common use http://en.wikipedia.org/wiki/German_tank_problem

[Richy_T]

I'm expecting a 20-30x difficulty rating, no meaningful increase in BTC/USD price, a halving of the block reward, and a ROI of around 6-8 months... Agree that people expecting to get all their $$$ back in a week are crazy but over the long term the BFL power consumption means it will continue to be paying for itself well into 2013.

Will

Are you counting the ROI from when you start mining or when you sent your money?

[cedivad]

The money vs performance ratio is better because their drivers suck and nobody would buy their products if they were even with Nvidia.  Did you ever notice if you install their driver on XP, it will load then reboot then tell you it relies on the .NET 2.0 framework and can't load anything in the driver's GUI without it?  Why would they write such big name software from such a wealthy company in freaking .NET?!  I'm a .NET programmer and I even think that was a stupid choice.

Then you've got crashes, crashes, and more crashes in any game that's been newly released followed by the game manufacturers advising you to turn off XXXX if you have an AMD card.  Nvidia doesn't seem to ever have that problem.  Then you've got shadows and lighting drawing incorrectly in 1 game and when they fix it, they break 5 other games.

I also couldn't possibly care any less about Linux support

I think that they use .net for something different than the core drivers.

Most probably the GUI of the AMD control center or something like that.

It would make sense to me.

Np at all with drivers my side, i must be lucky.

[Desolator]

yeah, .NET wouldn't be idea for the driver operations  Although they obviously use DirectX lol.  But yeah, it's the freaking interface!  To just show a window on a screen to millions of customers, they picked the least likely to work method and then don't even include the framework installer itself in the driver or a downloader for it!  That's like a $100,000 company mistake not a $1 billion+ company mistake.

[scrybe]

yeah, .NET wouldn't be idea for the driver operations  Although they obviously use DirectX lol.  But yeah, it's the freaking interface!  To just show a window on a screen to millions of customers, they picked the least likely to work method and then don't even include the framework installer itself in the driver or a downloader for it!  That's like a $100,000 company mistake not a $1 billion+ company mistake.

Cough. Oracle. Cough.

Nobody is immune from bad installer decisions, larger companies just have more inertia, and tend to get stuck longer.

[Korbman]

In 6-12 months those ASIC devices will be replaced with better equipment.
During that time they will not lower their Price and here is why.

Reason 1. Have a product to offer that is better then your competitor makes you money.
Reason 2. During the time of development they will have to fund that research.
Reason 3. If you lower the price of the devices your existing customers bottom line will be effected.

Solution 1. Reveal a 2nd generation device before your competitor.
Solution 2. If it makes business sense offer full value trade-in on  first gen pre-order devices.
Solution 3. If you happen to lower your prices. Offer trade-in options will less credit towards 2nd generation pre-order devices.

IMHO To have a successful business and or project you should adopt a policy that will never burn your bridges with your customers.... Ever...

Succinctly put and overall quite right.
I actually remember a quote from one of my business classes later in college, and it comes from Sam Walton: "There is only one boss. The customer. And he can fire everybody in the company from the chairman on down, simply by spending his money somewhere else."
In other words, keep your customers happy and they'll keep coming back to you. BFL had some trouble with this to start, but given they have a fair amount of competition this time around I don't see why they wouldn't want to make us happy. Otherwise I'll go buy from bASIC

Oh, and:

The customer should never see the order# (or at least the row number in the database). They understood this even back before computers were in common use http://en.wikipedia.org/wiki/German_tank_problem

LOL yay for vastly overestimating quantities!

[scrybe]

If you go through the order process and never fwd the bitcoins or payment the order number still goes up by 1.

Ohhhh, they should clean that up, lol.

The customer should never see the order# (or at least the row number in the database). They understood this even back before computers were in common use http://en.wikipedia.org/wiki/German_tank_problem

You need a unique reference to run a business, the customer needs that reference as well. The best they could do is obfuscate it (with double SHA2?) or something, but why would makers of web-store software even care about that in a general case? Outside this fishbowl community attempting to decode order numbers is generally a futile exercise, so why expend the effort on the feature?

[Richy_T]

You need a unique reference to run a business, the customer needs that reference as well. The best they could do is obfuscate it (with double SHA2?) or something, but why would makers of web-store software even care about that in a general case? Outside this fishbowl community attempting to decode order numbers is generally a futile exercise, so why expend the effort on the feature?

Yes, you do. But the rowid of the database is not the way to do it. A better way is to choose an arbitrarily large number space, choose a number from it randomly and check for uniqueness. There are other methods too. At the least, they should not be easily guessable.

[SgtSpike]

I'm expecting a 20-30x difficulty rating, no meaningful increase in BTC/USD price, a halving of the block reward, and a ROI of around 6-8 months... Agree that people expecting to get all their $$$ back in a week are crazy but over the long term the BFL power consumption means it will continue to be paying for itself well into 2013.

Will

The people expecting to get all their $$$ back in a week are the ones that had preorders in on the first day, and will (likely) be receiving the first ASICs out.  And those people very likely will have a 100% ROI in around a week, provided the BTC price doesn't take a sudden dive.

I'll let you know. 

[Desolator]

Uhhhh maybe you should read BFL's shipping policy before saying that I had assumed everyone reading this post had done that but that does not appear to be the case 

Because of the potential for massive disruption that our ASIC units may cause to the mining ecosystem, we are instituting a 1/3 shipping method for our first shipment of products to individuals.  So that no single person or handful of people are able to gain a material advantage over any other, we will ship all our available units at the same time to as many customers as we can.  This means we will likely delay shipping on our first units until we have accumulated enough stock to satisfy a significant portion of our initial orders.  This ensures that as many people as possible who are early adopters are equally and fairly treated when receiving our product. To that end we will be shipping our first orders as such: 1/3 of each product line (Jalapeno, Single SC and Minirig SC) will be shipping to new orders, 1/3 to upgrade orders and the final 1/3 will be shipped to a random selection of orders from the first two categories (both upgrade and new customers)  received in the first month of orders (From 23 June 2012 to 22 July 2012).

That unfortunately results in an almost 100% chance that you won't get 100% of your order during the initial shipment and also that if you didn't place an order before July 2012, you're definitely not getting 100% of your order (if you ordered more than 1)

[meebs]

You need a unique reference to run a business, the customer needs that reference as well. The best they could do is obfuscate it (with double SHA2?) or something, but why would makers of web-store software even care about that in a general case? Outside this fishbowl community attempting to decode order numbers is generally a futile exercise, so why expend the effort on the feature?

Yes, you do. But the rowid of the database is not the way to do it. A better way is to choose an arbitrarily large number space, choose a number from it randomly and check for uniqueness. There are other methods too. At the least, they should not be easily guessable.

Unfortunately in SOME form of the business system there will need to be an orderly ascending unique identifier for each transcation. Typically this is done at the invoice # level where each new invoice is +1 above the previous, regardless or how/what/why. This is needed for auditing purposes.

That said.. for a sales order #, there is likely a lot more freedom with that, but i would imagine it would still need SOME kind of rhyme or reason vs just being totally random.

[johnyj]

The fact is:

Even BTC price drop to 1$, there are plenty of people waiting there to pick all of them up
Even difficulty drop to very low, there are plenty of people waiting there to mine more coins

Since all of them know that the total amount of BTC is limited

[FLHippy]

yeah, .NET wouldn't be idea for the driver operations  Although they obviously use DirectX lol.  But yeah, it's the freaking interface!  To just show a window on a screen to millions of customers, they picked the least likely to work method and then don't even include the framework installer itself in the driver or a downloader for it!  That's like a $100,000 company mistake not a $1 billion+ company mistake.

Cough. Oracle. Cough.

Nobody is immune from bad installer decisions, larger companies just have more inertia, and tend to get stuck longer.

C'mon haven't you people ever worked for a real company?

As a lead developer you provide a project plan to do the project right.

As a mid level manager you find ways to shave time and dollars off of the project plan.

Developers complain "but it will suck"

They give the installer task to the intern who is sucking the mid level managers cock. She can't code but she works late and produces poor quality rubber band and duct tape shit spaghetti code  quickly.

The boss's boss gives the boss a bonus for coming in under budget.

Neither the boss, nor the boss's boss will ever (EVER) actually look at the software.

The support team deals with the fallout and hates the developers for being so lazy.

It's how it works.. When you've been doing this for 22 years like I have you just roll with it. You don't get upset when projects you put your life and soul into get tossed in the trash because of the color of a button (but we can change that!) just so some manager can get credit for having produced a revolutionary new software product. You become jaded, you've already written everything ever so you become a cut and paste programmer while reading news on the web and running your bitcoin side business from your relatively high paid senior developer job.

All is right in the world...

[Desolator]

But at my company, we have 1 programmer: me So all the code works flawlessly and I don't have to deal with incompetent morons that are in programming for the money instead of due to their intelligence level.

During college I had to work on a huge car rental program and my partners in my group were a guy that was dropping out the next semester and a 55 year old lady being retrained by her work.  It got a score above 100% because I wrote 90% of it plus the bonus extra hard part

[Richy_T]

It's how it works.. When you've been doing this for 22 years like I have you just roll with it. You don't get upset when projects you put your life and soul into get tossed in the trash because of the color of a button (but we can change that!) just so some manager can get credit for having produced a revolutionary new software product. You become jaded, you've already written everything ever so you become a cut and paste programmer while reading news on the web and running your bitcoin side business from your relatively high paid senior developer job.

Someone's hacked into my autobiography...

[Richy_T]

You need a unique reference to run a business, the customer needs that reference as well. The best they could do is obfuscate it (with double SHA2?) or something, but why would makers of web-store software even care about that in a general case? Outside this fishbowl community attempting to decode order numbers is generally a futile exercise, so why expend the effort on the feature?

Yes, you do. But the rowid of the database is not the way to do it. A better way is to choose an arbitrarily large number space, choose a number from it randomly and check for uniqueness. There are other methods too. At the least, they should not be easily guessable.

Unfortunately in SOME form of the business system there will need to be an orderly ascending unique identifier for each transcation. Typically this is done at the invoice # level where each new invoice is +1 above the previous, regardless or how/what/why. This is needed for auditing purposes.

That said.. for a sales order #, there is likely a lot more freedom with that, but i would imagine it would still need SOME kind of rhyme or reason vs just being totally random.

Given that the first site I checked, Newegg appears to have such a system for their order numbers, I'll concede this (not sure what their invoice numbers are doing though). Thinking back, when I worked on an e-commerce site back in the late 90s, we used an increasing-by-1 system but just started with an arbitrarily large number.

[scrybe]

yeah, .NET wouldn't be idea for the driver operations  Although they obviously use DirectX lol.  But yeah, it's the freaking interface!  To just show a window on a screen to millions of customers, they picked the least likely to work method and then don't even include the framework installer itself in the driver or a downloader for it!  That's like a $100,000 company mistake not a $1 billion+ company mistake.

Cough. Oracle. Cough.

Nobody is immune from bad installer decisions, larger companies just have more inertia, and tend to get stuck longer.

C'mon haven't you people ever worked for a real company?

As a lead developer you provide a project plan to do the project right.

As a mid level manager you find ways to shave time and dollars off of the project plan.

Developers complain "but it will suck"

They give the installer task to the intern who is sucking the mid level managers cock. She can't code but she works late and produces poor quality rubber band and duct tape shit spaghetti code  quickly.

The boss's boss gives the boss a bonus for coming in under budget.

Neither the boss, nor the boss's boss will ever (EVER) actually look at the software.

The support team deals with the fallout and hates the developers for being so lazy.

It's how it works.. When you've been doing this for 22 years like I have you just roll with it. You don't get upset when projects you put your life and soul into get tossed in the trash because of the color of a button (but we can change that!) just so some manager can get credit for having produced a revolutionary new software product. You become jaded, you've already written everything ever so you become a cut and paste programmer while reading news on the web and running your bitcoin side business from your relatively high paid senior developer job.

All is right in the world...

Exactly, what manager is going to defend that feature azhead of their bonus? If you really want the feature to stay in, they have to really want it too.

Software never makes it to ideal, someone always releases it...

[VeeMiner]

Yep, you are all screwed. Everyone cancel your orders so I can make a lot of money.

Thanks  

+1