Quote from: MonsterZero on February 11, 2013, 09:15:53 PM

Quote from: PuertoLibre on February 11, 2013, 09:09:29 PM

Quote from: MonsterZero on February 11, 2013, 08:32:26 PM

Quote from: PuertoLibre on February 11, 2013, 08:27:47 PM

Quote from: Evan on February 11, 2013, 08:15:58 PM

Quote from: K1773R on February 09, 2013, 08:45:53 PM

Quote from: Turbor on February 09, 2013, 09:07:24 AM

About time

About DAMN time

So if i bought one today, I am waiting untill November? bull shit... they better drop that price by $300 if they think I am waiting 7 months for delivery of the products.

MFG costs should go down as time continues, or the Machines should have considerably more hashing power, or Power efficiency a year later.

If you start off with a tiny 65nm process that is extremely densely packed...

You can't expect that a 45nm or 32nm product to cost the same or less. It has to cost more. The only way to bring down the price is at the cost of profit.

http://en.wikipedia.org/wiki/Die_shrink

Die shrinks are popular among semiconductor companies, such as Intel, AMD (including the former ATI), NVIDIA, and Samsung for enriching their product lines. Examples in the 2000s include the codenamed Cedar Mill Pentium 4 processors (from 90 nm CMOS to 65 nm CMOS) and Penryn Core 2 processors (from 65 nm CMOS to 45 nm CMOS), the codenamed Brisbane Athlon 64 X2 processors (from 90 nm SOI to 65 nm SOI), and various generations of GPUs from both ATI and NVIDIA. In January 2010, Intel released Clarkdale Core i5 and Core i7 processors fabricated with a 32 nm process, down from a previous 45 nm process used in older iterations of the Nehalem processor microarchitecture.
Die shrinks are beneficial to end-users as shrinking a die reduces the current used by each transistor switching on or off in semiconductor devices while maintaining the same clock frequency of a chip, making a product with less power consumption (and thus less heat production), increased clock rate headroom, and lower prices.

The lower prices are due to mass volume production of other parts. Keep in mind BFL and Avalon are selling entire units. NOT single chips like those you cited.

They (AMD, Intel etc) also sell hundreds of millions of units (chips) so the NRE costs are a tiny fraction of the total cost. Scale <--- makes the price come down.

With ASIC's that is not the case. It is one batch for a few hundred people and only a number of wafers. (Up to a [low] multiple of 10,000 chips.)

(Note: For example, BFL only did 6 wafers for around ~6000~ ASIC chips in their first batch)

The cases and the rest of the components do not come down in price unless they establish a long term contract for the same parts. (in other words, not a single run)

This is basic economics of scale. It should be extremely obvious.

Interesting change of subject. You were speaking about nm scale production costs.
You said "You can't expect that a 45nm or 32nm product to cost the same or less. It has to cost more. "
Which is incorrect. It should be extremely obvious.

Wow,

So then answer this: Why didn't BFL go for 45nm technology from the start? Any guesses?

Edit: I think I understand your confusion. You think I am talking about the ASIC chips only and not the entire unit.

NO, I am referring to the entire unit. No ASIC company to date sells their chips like an AMD or Intel. Adding NRE costs to a smaller nm does increase the number of chips. But if done wrong usually results in less viable chips due to tiny defects.

AMD or Intel use binning and scale to get around that issue. As far as I know, I dont' see BFL or Avalon using binning techniques.