Topic: [ANN] Bitfury ASIC sales in EU and Europe - page 6. (Read 250419 times)

I think the only reason he "ROI'd" is because of the run up in Nov 2013.  He would have likely done waaaaayy better buying BTC In Sept and holding.

It wasn't my intention to claim such. I'd love to buy bitfury stuff again, but we are not in 2013 anymore and the chip has to be really incredible to make sense to sell it to the public.

Seriously guys, use some sense. They posted the first real life test results from an ENGINEERING SAMPLE and suddenly everyone is claiming they just deployed 100PH of them.

Seriously guys, use some sense. They posted the first real life test results from an ENGINEERING SAMPLE and suddenly everyone is claiming they just deployed 100PH of them.

After rigorous testing, the new BitFury Chip has demonstrated outstanding computing capabilities. The design target was 40 gigahash per second with power efficiency of 0.06 joules per gigahash. On average, the measured power efficiency of tested engineering samples of the new BitFury Chip ranges from 0.055 joules per gigahash to 0.07 joules per gigahash, thus making the BitFury 16nm ASIC Chip the industry leader in energy efficiency.

In addition to its outstanding power efficiency, BitFury’s 16nm ASIC Chip can run with as low as 0.35V power supply voltage. The static part of the Chip can be functional at as low as 0.28V power supply voltage. This level of productivity was previously unattainable with existing silicon devices and existing silicon design methodologies, with transistors working in active mode, not slow sub-threshold. J / Gh metric in working modes starting from 55 Gh/s up to 180 Gh/s follows an almost linear relationship of 0.0011, while at 40 – 55 Gh/s measured slope converges to plateau.

Every BitFury 16nm Chip delivers a minimum of 100 gigahash per second of computing power. On average, the BitFury 16nm ASIC can compute in the range of as high as 140 gigahash per second using air cooling, and up to 184 gigahash per second using immersion cooling, which makes the BitFury custom-made silicon an absolute leader in computational efficiency per unit of silicon area.

*picture of hashrate graph*.

pics or it didn't happen  

The chip is crazy and amazing.

I think people are more interested in all the non chip parts of this, power consumption can only sell people so far.

• What will the estimated MOQ be? $1k, $10k, $100k etc.
• Please don't announce preorders...
• How much engagement we will get with the community which isn't done through a PR agency?
• How will these be sold? Please don't say system integrators and please don't say by resellers, both of which have been a detriment to end users.
• Are they going to be priced at remotely viable levels? I'm not one to comment on profitability usually but lets at least get in the ballpark.
• When is Soon^TM, cutting through the PR when are we actually expecting products?
• Who is going to support sales and support? Bitmain currently has droves of people doing just that and I can't see the Head of Product Development sticking about to answer 200 emails a day.
• Will you finally release or showcase previous designs (BF3500 etc) outside of NDA? It adds credibility and trust to consumers knowing what previous machines looked like. The last thing we saw were H-boards, which while sufficient at the time are a pretty outdated design.

I feel this - like many other previous announcements - are the diamond ring at the proposal. I just hope we don't get mistreated in the rest of the marriage.

 We agree again time will tell how good it is  to buy them.

I was just about to express my doubts, about the technology scaling factor and the achieveable frequrencies around 0.4V with 16nm FinFETs, but your EDIT sounds much more feasible.

According to my crystal ball from last night and coffee stains from this morning ... some chip specs:
1. Minimum voltage for stable operation should be 0.4V and 10A for about 60Gh/s
2. Air cooled nominal power options: 0.45V and 10A  OR  0.5V and 14A  OR  0.6V and 12A with lower freq, if the internal oscillator is programmable and voltage dependant as in the 55nm chip
3. Immersion cooled power 0.65V and 20A

Some wild guesses and speculations about the internal structure:
Let's assume 4096 cores, because on the same die size at 16nm you can fit 4 times more. Using the same rolling cores (61+4 cicles) from the 55nm chip this gives 1GHz min working freq.
Then the nominal working frequency should be ~2GHz and up to 2.9GHz with immersion cooling

EDIT: Actually you can fit 16 times more (16*864=13824) and the frequency difference above is too big, so some other options:
 with 8192 cores the numbers are coming down to 500MHz min, 1GHz nominal and 1.45GHz ... still too big frequency difference
 with 10240 cores: 400MHz min, 800MHz nominal and 1.1GHz
 with 13824 cores: ~300MHz min, ~600MHz nominal and ~860Mhz ... seems normal, so that is more likely the case

The design target was 40 gigahash per second with power efficiency of 0.06 joules per gigahash ...
... BitFury’s 16nm ASIC Chip can run with as low as 0.35V power supply voltage ...
... from 55 Gh/s up to 180 Gh/s follows an almost linear relationship of 0.0011, while at 40 – 55 Gh/s measured slope converges to plateau ...
... Every BitFury 16nm Chip delivers a minimum of 100 gigahash per second of computing power. On average, the BitFury 16nm ASIC can compute in the range of as high as 140 gigahash per second using air cooling, and up to 184 gigahash per second using immersion cooling

I think the only reason he "ROI'd" is because of the run up in Nov 2013.  He would have likely done waaaaayy better buying BTC In Sept and holding.

Quite different times when ASICs fist came out back then, but that is good to hear they ROI'd so fast the first go round!

But in the same time first unit did an excellent ROI.
My 25 GHS starter kit delivered 1-st of Sept 2013 made 4-5 times ROI by the end of the year

Also bought some chips and used assembly service for them 1-st gen
Got 3-4 times ROI

So  in fact all my bitfury based gear made ROI