Eyes to the horizon: A bitcoin ASIC project will be announced in June

[aiwill]

Also, companies like KnCMiner claim 28nm ASIC.  I recall another person (i cannot recall the name) mentioning they were in the works for creating 28nm asics, but would only sell to companies (like a real semiconductor company usually does..)

Also, Bitfury is 28nm i think?

Seems like this is a frequent question.  I d like to explain it minutely

1. In the present case (hundreds T size) , 28nm is totally unrealistic, it sounds like: pay millions more to save one dollar for electricity charges. With a rough calculation, 28nm would probably get a real advantage after hash rate achieve 50k+T

2. The thought that 28nm&45nm is "better" than 130nm is purely amateur, which nm number to choose depends on many factors, also,good optimizations in both FPGA and ASIC increase the function of chips so much,    which means that, even for 130nm/110nm, there are still  several aspects exist to improve.  

3. I have no idea about what is the real plan of bitfuty/KNC. I am not saying that theirs are fake, but I see no reason to develop 28nm at the moment,which is, extreme expensive/sounds pretty cool/highest risk and lowest return.

[1715295579]

1715295579

[1715295579]

1715295579

[Franktank]

+1 to Mode A

[PeZ]

Instead of letting the rich get richer by selling your equipment, go for mode A and let everyone in on it.

[notlist3d]

+1 for honest reply on 28 nm.   Love to see some honesty.

[ninjaboon]

will be watching although bitcoin mining is really getting competitive.

[tempestb]

So then how is this going to be better than what Asicminer offers now?  If you have a similar 10gh board.  You'll need, what, ten thousand boards to reach 100th?  You're really going to build that many?  

[AvenG]

How about decentralizing the hashpower by selling part of your chips and rigs? I cannot trust in these kind of huge mining farms business model.
Also, model A sounds good even though you're not bulletproof against natural disasters. You could spread your ASICs in multiple locations all around China or you could sell part of them to your worldwide investors while securing the 100T power.

[cchan]

I prefer Mode A.  

[Wayne_Chang]

I prefer to mode A. I think in this mode the whole project will start to run in a short time. As now time is a critical factor to get more profit of mining BTC, this mode will be better.
I am looking closely to buy some shares. 

[ecliptic]

How about decentralizing the hashpower by selling part of your chips and rigs? I cannot trust in these kind of huge mining farms business model.
Also, model A sounds good even though you're not bulletproof against natural disasters. You could spread your ASICs in multiple locations all around China or you could sell part of them to your worldwide investors while securing the 100T power.

to be fair, Avalon is already doing that, and their chips are nearly identical in terms of size, hashing power, wattage, etc.

[GGGGG]

赞! Here's another vote for Mode A as well.

Are you guys located in Beijing?

[HeRetiK]

Voting for Mode A and keeping an eye on this.

[punin]

Also, companies like KnCMiner claim 28nm ASIC.  I recall another person (i cannot recall the name) mentioning they were in the works for creating 28nm asics, but would only sell to companies (like a real semiconductor company usually does..)

Also, Bitfury is 28nm i think?

Seems like this is a frequent question.  I d like to explain it minutely

1. In the present case (hundreds T size) , 28nm is totally unrealistic, it sounds like: pay millions more to save one dollar for electricity charges. With a rough calculation, 28nm would probably get a real advantage after hash rate achieve 50k+T

2. The thought that 28nm&45nm is "better" than 130nm is purely amateur, which nm number to choose depends on many factors, also,good optimizations in both FPGA and ASIC increase the function of chips so much,    which means that, even for 130nm/110nm, there are still  several aspects exist to improve.  

3. I have no idea about what is the real plan of bitfuty/KNC. I am not saying that theirs are fake, but I see no reason to develop 28nm at the moment,which is, extreme expensive/sounds pretty cool/highest risk and lowest return.

Bitfury's chip is 65nm that has been optically scaled to 55nm.

[Bicknellski]

A project of ASIC bitcoin mining is ongoing by our group, named btc-garden (www.btcgarden.com).  At the moment, we are at the stage of TAPE OUT.

We have started our project since Feb.2013.Also, according to our timetable, we will have around 100T hashrate online on this August or September as our best situation. The following paragraphs are written to give out a brief introduction of this item. A relative detailed version will be posted on this board & our website at the beginning of June, 2013. Any questions or advices are welcomed in this thread.

A brief introduction

1.   About Us
BTCMAN (WWW.BTCMAN.COM, portals & information site in Chinese,with 15k registered members by far) is a full time bitcoin-related team from china; we currently have a decent PV of our forum & the main site.
 

In terms of the project of BTC GARDEN, all of our core members were graduated from CAS (Chinese Academy of Sciences), after graduation, some of them stayed there to do research works till now, while others make a living at Intel or Nvidia, I.e. we have both R&D Personnels from institute and practical engineers from mature industrial system. From a technical perspective, the difficulty of this project is really easier than many of our previous ASIC projects.
List of Us
ID：W.J, team leader, 8 years experience in system design.
ID：D.M , 3 years Intel experience in VLSI design( front-end and back-end ),5 years
Experience in system design and software design. In charge of ASIC design and software design.
ID: Xrange.G, 8 years Spreadtrum experience in VLSI design (front-end). In charge of RTL design, RTL verification.
ID: Kevin.S, 12 years HuaWei experience in VLSI design (front-end and back-end). In charge of FPGA verification, synthesis.
ID: Y.S, 4 years NIVIDA experience in VLSI design (back-end). In charge of physical design.

2.   Tech Exhibition
a)   Project Milestone
i.   March 15th: We also signed the confidential contract with the IC manufacture and got the library for synthesis and layout.
ii.   April 10th: Our RTL design, optimization and simulation were finished. We have some data to predict the specification of actual chips.
 

iii.   May 16th: layout Version 1 finished.
 
 

b)   ASIC brief introduction
SMIC 0.13um;
Core Voltage: 1.2V;
I/O: Voltage: 3.3V;
Core Frequencey: 400MH/s;( @vdd 1.25V);
Number of Pads; 64;
Package lqfp64;

3.   The Business Idea

Initially we have the following two possible modes. The final decision will depend on several factors like the feedbacks from BTC world, the exchange rate of BTC in JUNE…etc.

Mode A: we spend our own money in creating the first 100T, DIY everything without any requests from public: imagine everything go successfully, we shall be able to mine new blocks with 100T hashrate after that , we will then launch a bitcoin-stock for IPO, the stock can trade freely on our site later. Also, all of shareholders will get lifelong dividends every several weeks, which means that no matter how much more new hashrate be added by us, shareholders will get corresponding share dividends with 100% profit. (I.E., we will allocate all profit without any retained earnings in every term.)


Mode B: we simply sell out our hashrate to public, with a price around 1000-1200 BTC/T/first year(50G as minimum unit, there will be around 100-200T for sale.)We will get in charge with everything. All buyers will get their income weekly.  Within this mode, we probably open a two-months-long period for people to buy and pay , then close it up after that time/sold out. Mode B will be acceptable only if we can sell out at least 40T.

In terms of the mining method, we will prefer solo to any pools because of the instability of them. (The point that pools can help preventing deceptive practices is inexplicable, transparency can never be guaranteed by this )


    (BTW: Most likely,we will also try single mining hardwares like BFL`s products or avalon`s chips in next generation)


4.   Q & A

Q:Do you guys have any relationship with ASICMINER or AVALON?

A: No, but we do have a similar background. We also had some contacts with them through either IM tools or in person, although some of their thoughts / behaviours are hardly accepted by us.

Q: how to guarantee the transparency ? I know that a 100% provable openness is impossible in BTC world,but what can you offer as your best?

A: A daily video will be upload every day to show out how much we have digged out. It will be better than pure pictures or photos which can easily be photoshoped. We may adopt a lot of methods like this to reduce the probability we do anything dirty.

Q: Is there a list to show out all the solutions of those possible accidents? For instance, chips broken within first year; unexpected huge hashrate influent our income so much;equipments in the machinery rooms get stolen; Website get hacked;how long time is needed for deploying the first 100T;how to define the”first year”, if there are some offline-days in the first year, is any kind of compensation/refund offered…etc.

A:As we metioned on the 3rd part, we still haven’t decided which mode to choose yet, if we finally prefer Mode B to A, then all kinds of questions may be posted on somewhere at our website and will be replied asap.



Again,any questions or advices are welcomed in this thread.

Q: Chips for sale?
A: (Please say yes)

[Rampion]

Seeing the pricing you propose for 1TH in model B (1,000-1,200BTC/TH/year), I'm afraid your IPO price will be too expensive. I would take into account that Bitfountain's IPO price was 0.1BTC per share, for an expected deployment of +250TH. If I do not recall it wrong, each share is 1/400,000 of the total profit/hashrate, which means that their price for TH was 160BTC, not a hefty 1,000-1,200BTC price tag.

[thatbluedude]

Seeing the pricing you propose for 1TH in model B (1,000-1,200BTC/TH/year), I'm afraid your IPO price will be too expensive. I would take into account that Bitfountain's IPO price was 0.1BTC per share, for an expected deployment of +250TH. If I do not recall it wrong, each share is 1/400,000 of the total profit/hashrate, which means that their price for TH was 160BTC, not a hefty 1,000-1,200BTC price tag.

as I understand it they will have working asics at ipo, asicminer didn't.
significantly smaller leap of faith -> higher price

[jpyao78]

interested.

[GodHatesFigs]

Seeing the pricing you propose for 1TH in model B (1,000-1,200BTC/TH/year), I'm afraid your IPO price will be too expensive. I would take into account that Bitfountain's IPO price was 0.1BTC per share, for an expected deployment of +250TH. If I do not recall it wrong, each share is 1/400,000 of the total profit/hashrate, which means that their price for TH was 160BTC, not a hefty 1,000-1,200BTC price tag.

Even at today's difficulty, a TH would take about a year to earn 1200BTC. Ridiculously overpriced, especially as that price only leases you the hash rate for a year.

[JimiQ84]

Seeing the pricing you propose for 1TH in model B (1,000-1,200BTC/TH/year), I'm afraid your IPO price will be too expensive. I would take into account that Bitfountain's IPO price was 0.1BTC per share, for an expected deployment of +250TH. If I do not recall it wrong, each share is 1/400,000 of the total profit/hashrate, which means that their price for TH was 160BTC, not a hefty 1,000-1,200BTC price tag.

Even at today's difficulty, a TH would take about a year to earn 1200BTC. Ridiculously overpriced, especially as that price only leases you the hash rate for a year.

At today's difficulty, 1 TH/s will make 1200 BTC in 27 days. (without electricity costs). You know, 1200 BTC per 1 TH/s is 1.2 BTC per 1 GH/s. Similar price as Avalon #3

[tom_o]

Seeing the pricing you propose for 1TH in model B (1,000-1,200BTC/TH/year), I'm afraid your IPO price will be too expensive. I would take into account that Bitfountain's IPO price was 0.1BTC per share, for an expected deployment of +250TH. If I do not recall it wrong, each share is 1/400,000 of the total profit/hashrate, which means that their price for TH was 160BTC, not a hefty 1,000-1,200BTC price tag.

Even at today's difficulty, a TH would take about a year to earn 1200BTC. Ridiculously overpriced, especially as that price only leases you the hash rate for a year.

At today's difficulty, 1 TH/s will make 1200 BTC in 27 days. (without electricity costs). You know, 1200 BTC per 1 TH/s is 1.2 BTC per 1 GH/s. Similar price as Avalon #3

True but with that you actually own hardware that can be resold etc, with this you just wait to hopefully get paid more than you put in.