[BitCentury] Metabank 120Gh 65nm Pre-Order Proxy [CLOSED]

[Foofighter]

How's this for more Bitfury news...

Bitfury started an open bet here: http://bitbet.us/bet/450/bitfurys-asic-will-work-with-power-1/

...and placed 99 BTC of his own money on the YES side. The Yes side meaning that he if he didn't deliver less than 1 W per GH/s per chip by June 15th, 2013, he'd lose his coins to the opposing betters.

Well June 15th came and went, and he missed the deadline by a few days, so he lost the bet from a time perspective. Now a few days later his chip is proven to be working and supposedly exceeding the power/Gh numbers, and so he is actually paying out everyone else who bet on the YES side who believed in him - something he never had to do.

More details here:
https://bitcointalk.org/index.php?topic=228677.msg2516092#msg2516092

Genius + high integrity = Awesome! 

yeah, also saw that, really a nice guy!

[digitalmagus]

Hi guys,
It's been a few days since we've provided some updates. This is mostly because we've been busy dealing with some administrative work in the background. FYI, we have not yet received our alpha chips. It looks like the first ones went to individuals primarily involved in producing PCBs for Metabank and similar parties that were investors in Bitfury's chip manufacturing, so this is not a bad thing. There is apparently a larger set of chips already in Europe that will be shipped out to all the remaining selected testers such as our organization; however whilst we have no exact ETA we were told 2 weeks, about 1 week ago. So let's see what happens there. In the mean time, one of the Bitfury chips was sent to zeptobars.ru, which is a site that apparently analyzes chips under high zoom optical microscopes.

Here's the relevant eyecandy:

QFN packaging - execellent for thermal performance and low inductance connection to leads:


Chip without the packaging - tiny! This plus low voltage operations, should make for some awesome USB hashers one day.


If we correctly understood some of the Russian postings, bitfury's chip will have over 700 cores (crazy). This is a good thing in the sense that for example, with BFL chips there's 16 cores (avalon I think has 1? Correct me if I'm wrong). If during the manufacturing process dust or other particles get in the die's surface, it could cause a core malfunction. This is why on some BFL chips, some people are saying that one BFL device performs slightly slower than another, because 1 or 2 cores may have been damaged. When you have 700+ cores, the hashing rate differential will be less noticeable. To be clear, this picture does not show identifiable cores.


Nano art on the chip. Apparently Bitfury had some spare time on his hands ;-) Surprised it wasn't a bitcoin symbol or the logo from his FPGA website: http://bitfury.org.


Etching materialization


Repetitive pattern of hash calculation blocks:


55nm is at least 5 times smaller than what we can see in optical microscope, so not all details are visible

[digitalmagus]

Hi guys,
We recently spoke with  Metabank and here's some information we'd like to pass on:

Q1: When will the machines be ready for delivery?
A1: In Autumn, around October.

Q2: We heard of at least 1 individual on Bitcoin forum that was selling Metabank devices after you announced Batch 1 was closed, can you explain this? Do you have any more units you can sell us?
A2:The batch is closed, we have not sold more equipment since we announce the closure.

Q3: Will your product be a bare circuitboard or will it ship in a case like avalon and BFL have.
A3: Most likely it will be in a case

Q4: Do you know if it will be dual voltage power supply? (110/220V). How many watts will the unit draw at the wall?
A4: Sorry, I don't know answers to either of those questions. We are not at that stage yet.

Q5: Can you explain if the interface from the computer will be USB or Ethernet?
A5: It most likely won't be connected to the computer, just the power plug. It's independent from the computer .

Q6: Yes, but how'd the computer intercommunicate with it? What's the interface?
A6: There'll be software for that

Q7: So it'll require special software to run from the computer? Which mining software? cgminer? something else?
A7: Yes. I can't answer questions like this yet. If I make a mistake, I'll leave you misinformed.

Q8: What is your estimate for when there might be a Batch 2? When is Bitfury getting Batch 2 chips from factory?
A8: I can't tell the exact dates at this point. Around the release of the Batch 1, we'll have understanding when the Batch 2 will be pre-ordered.

As we get more details over time, we'll do our best to pass them along. Right now, i'd say they're in very early stages of designing the PCBs, so obviously not all the details are worked out yet. PCB design can take a couple of weeks to a couple of months, depending on complexity, access to all the necessary equipment, engineering skill level, design flaws/re-starts, etc. Likewise for the chip assembly process on the PCBs, although that part can be done in about 5-10 days, depending on volume and availability of all the tiny components (aside from the ASICs).
Cheers,
Luis


PS. I'm a little behind on waitlist requests, but will try to get to those in the next couple of days. Thanks for your patience.

[digitalmagus]

Greetings,
Just another quick update. We can confirm that our Alpha chips are officially en route! ETA = Mon or Tues (depending on time zone).  In case you are wondering, the Bitfury alpha chips will allow us to:
a) Peer review and confirm the chip's functionality, hashing speed potential and power usage
b) Allow us to start thinking about possible PCB designs in the event Bitfury ever decides to sell chips like Avalon or BFL.

Have a great weekend everyone!
Luis

[fpgaminer]

We can confirm that our Alpha chips are officially en route! ETA = Mon or Tues (depending on time zone).

Very cool!  Do you have a prototype PCB ready for them, or are you dead bugging it?

[tacotime]

The reported efficiency numbers from the Russian subforum posts are almost unreal, it appears at low voltages he's hitting 0.40 W per GH/s.

[digitalmagus]

The reported efficiency numbers from the Russian subforum posts are almost unreal, it appears at low voltages he's hitting 0.40 W per GH/s.

Tacotime, what a pleasure to see you in our thread! This reminds me, I need to catch up on your MC2 Alt-coin thread (URL: https://bitcointalk.org/index.php?topic=169204.0). My initial impression was that it looked like a promising start to a real and improved alternative to bitcoin, as compared to all the other "clone-coins" out there (to use the polite terminology).

Very cool!  Do you have a prototype PCB ready for them, or are you dead bugging it?

@fpgaminer: I'll let Nigel comment on that, since he's the one that will be handling the alpha chip testing.

[OrphanedGland]

We can confirm that our Alpha chips are officially en route! ETA = Mon or Tues (depending on time zone).

Very cool!  Do you have a prototype PCB ready for them, or are you dead bugging it?

We had plans for various tests including dead bugging, but as we are receiving the chips later than originally anticipated, we are going straight to testing with a prototype PCB.  We are submitting our prototype PCB design for manufacturing today, lead time for that is approx 5-7 days.

[fpgaminer]

We had plans for various tests including dead bugging, but as we are receiving the chips later than originally anticipated, we are going straight to testing with a prototype PCB.  We are submitting our prototype PCB design for manufacturing today, lead time for that is approx 5-7 days.

Nice   Pics of the sexy, bare PCB would be eroticfun to look at!

[Xian01]

Wow, seems like this project is progressing very very nicely !

Any foreseeable trouble with shipping these units to the USA ? Interested in placing an order for one.

[Foofighter]

BITFURY chip hashing against 4 block erupter usbs!

[auto2nr1]

How fast is the BitFury chip?

[stripykitteh]

Q1: When will the machines be ready for delivery?
A1: In Autumn, around October.

Thank you for providing that update. Does anyone know why metabank are targeting the end of the range previously quoted (August-October)? I had the impression that the end of the range was for the circumstance that the chips would need a respin/redesign before mass production. The early testing suggests that bitfury beat his power/performance expectations, so that won't be necessary. I also note 100TH are getting their chips made up into boards over the next few weeks and are targetting the 3rd week of July for production.

Are metabank just being conservative? Is it because more work is required to get peak performance (~5GH/s) out of the chip for the finished machines?

[senseless]

How fast is the BitFury chip?

Right now it's been shown to operate at up to 4Gh/s. We'll know more once we begin testing the chips ourselves.

Q1: When will the machines be ready for delivery?
A1: In Autumn, around October.

Thank you for providing that update. Does anyone know why metabank are targeting the end of the range previously quoted (August-October)? I had the impression that the end of the range was for the circumstance that the chips would need a respin/redesign before mass production. The early testing suggests that bitfury beat his power/performance expectations, so that won't be necessary. I also note 100TH are getting their chips made up into boards over the next few weeks and are targetting the 3rd week of July for production.

Are metabank just being conservative? Is it because more work is required to get peak performance (~5GH/s) out of the chip for the finished machines?

I believe they are indeed just being conservative. As it's been shown with other asic suppliers assembly is a tedious process with many bits you need to be sure are in stock. If any of the bits are not in stock you're looking at a 2 week delay.

[TheSwede75]

I for one would not be surprised to see Bitfury ship as early as August based on the speed of the development process so far.

[OrphanedGland]

Hi All,

I just received some BitFury ASIC chips today.  Big thanks to BitFury for giving myself and the BitCentury team this opportunity.  We are currently waiting for our initial prototype PCBs to be manufactured.  I will keep you up to date with progress of the testing, but in the meantime check the following link for the pics: http://bitcentury.io/blog/bitfury-asics-have-arrived.

Regards,

Nigel / OrphanedGland

[OrphanedGland]

Hi All,

We received our prototype PCBs today, and have made very good progress with testing of the BitFury ASIC samples:

https://bitcentury.io/blog/initial-testing-of-bitfury-asic

Upcoming tests will include use of an external clock, and chaining of boards.

Regards,
Nigel / OrphanedGland

[Foofighter]

nice thank you nigel!
, any other news from metabank? Do we know how many chips metabank will use now? Or any other news from metabank?

regards

[fsb4000]

nice thank you nigel!
, any other news from metabank? Do we know how many chips metabank will use now? Or any other news from metabank?

regards

Last Metabank post : https://bitcointalk.org/index.php?topic=183368.msg2683984#msg2683984
"About the timing - everything is going according to plan. No artificial delays will not.
Remained the only risky step, the production of boards at the Russian factory. By this process, we carefully prepare in advance."

[senseless]

nice thank you nigel!
, any other news from metabank? Do we know how many chips metabank will use now? Or any other news from metabank?

regards

Much of the information is still undetermined at this time. The suppliers are still awaiting testing results (from bitfury, and alpha testers such as ourselves). Though preliminary results appear that 2-2.5 will be the operational range of the chip. It will likely be that the units will contain somewhere around 60 chips each. We will know more once we're able to conduct a full battery of tests on the chip (Our current results are only after less than 24 hours of testing and only on a single chip).