Bitcoin Forum

Hi there,

This is Canaan Creative from Beijing, working hard on the
upcoming 16nm SHA-256 mining chips for everybody!

If you are using Twitter: Canaan has started a Twitter feed
where we will post announcements regarding our upcoming
16nm chips over the next months.

https://twitter.com/canaanio

Please follow, stay tuned,
Canaan


working on a 16 nm  chip !!! ....

Price 2016-04-22   550 USD @1, 450 USD @10, 420 USD @50

Price is not bad for a 10 units. Still higher then the S7 but better build for sure.

you mean the  Avalon 6 ?
  

Isn't the A6 28 nm .

I only posted that because some have there doublets about Canaan Creative aka Avalon there one of the better company's to deal with all ways have been it's the cost of there miners that kind of turn a lot a way but they do have to date the best built miner used and new i have ever bought, i guess that justly the cost ...

If Avalon's working on something new, that is indeed good news.

I completely agree. Having Bitmain as the "sole" vendor is a poor situation. Yes, the price of the Avalon series has been kind non-competitive most of the time. It does keep Bitmain a bit more "honest" though, kind like Spondoolies did 12-18 months ago.

As unlikely as it is, I'd really like to see an affordable consumer miner out of Spondoolies. If nothing else, those guys knew how to treat a customer. But S5 sales broke the back of the SP20, the manufacturing cost of which brought to light a lot of weaknesses in their design philosophy.
At least in the Western world I think Avalon's gotten better with an English-speaking presence in BlockC; if that partnership existed a year earlier (and could provide reasonable domestic shipping rates) the Ava4.1 would have sold like crazy in North America. No contest the Avalon products are better built that Bitmains, but you do pay for it. Hopefully if they do come out with another competitive chip they at least try to make it available to third parties like they did for everyt chip before A3218.

Well I have yet to have bitmaintech compensate me for my batch 1 s-7.

I am running 8 or 9 avalon6's   so I look forward to a better avalon and will continue to support them.


Including helping with group buys etc.

*yawn* Nothing interesting to see here. Bitfury had an announcement like this from December and they don't have them in mass production. *yawn*

Yes absolutely, I was just checking the price on them ;-)

but that's great news for 16nm chips

Your point with these guys it's way different and what makes you think I'm not aware of bitfury's   BS post is how i see bitfury,


at least bitmain and Avalon when they say every one it's been everyone so far you can count on that and they release it on the dates they give us etc etc ....in other words they come thu .

I made this post more as info ..

my bad good point I'm even temped to buy one or two in a few weeks,

I'm not hoping for bitfury anymore. but will keep up hope that sidehack gets chips some place i have to many miners that need to be upgraded and parts just for those boards.

It is good news really for home/hobby miner.  Chances are bitmain is working on one to just has not made public.  But having Bitmain and Avalon at least makes it where not just 1 consumer model.

I really want to see specs and price.  They have made good miners in past (4.1 was one of my favorite miners of any due to how quiet it was).   But overall good news, hopefully they follow up with more.

I want  at least one more 4.1 for fun my power cost are about to go down wouldn't say a lot but enough to run a few 4.1's and get some thing that's about all i can afford right now unless i go into what little coin i have.


 i do hope there 16 nm chips they sell us are good i know it won't be cheap, all we can do is hope there .....

It was actually Spondoolies who started the price war, they just wanted to sell them at any cost in order to gain cash for the next gen. That just so happened to be far below what they could possibly have made them for, and above what Bitmain was manufacturing S5s for.


I'm not so sure, its price ceiling was above that of the S5 such that if there was sufficient competition that Bitmain could have squeezed them out. Avalon designs products how they envision they should (think Asian Spodoolies) where as Bitmain has become know for using the absolute 'efficiency' of design. Both are useful and important for a marketplace, both are different.


Its highly likely on both counts. Bitmain's [28nm] chip isn't astronomically good and seeing as Bitfury's [16nm] has gone walkies then the bar hasn't been set that high.

I never speculated on who started the price war. Just that Spondoolies lost, because the S5 ended up selling so much cheaper than an SP20. The SP20 was such an ungodly complex machine, I wouldn't be surprised if just the parts cost was higher than the S5's sales price by the end. Their design greatly increased the features and flexibility compared to the S5, but they crippled themselves on efficiency and cost-effectiveness.

Most of the people I heard complain about wanting to buy the 4.1 didn't complain about the price so much as the shipping cost from China. Maybe we were reading different posts, but the consensus I gathered was if Avalon had a North America distributor that didn't add on the order of $100 per unit to ship, there would have been a lot of sales. I'd have probably bought one, and I didn't buy anything last gen.

If they did a deal with blockC  for the avalon 4.1 it would have sold well.

forgetting price  I would rank the 4.1 as the best of the three.

It was the only one of the three I considered buying at the time.

4.1 was almost silent it was a truly great miner, I loved it so much I started with 1 and ended up getting 2 more for a set of 3.  Price was a little harder to justify, but I liked it being silent at that time I was mining in room's in my house before my mining area.  So was worth it to me.

The good new's is they seems to have kept up more with Bitmain pricing with Avalon 6 then then did with Avalon 4.1.  So assuming a new machine and they compete with bitmain it should be in the same ball park (hopefully).

But it's really hard to say as so much is unknown.

I hope Avalon has been working on a 16nm chip longer than a month or two, or they're going to be VERY late to market with it.

 The announcement they made that they weren't GOING to work on one for a while as of the release of the A6 was a very bad sign for their timing......

Ah  you can't believe them as Avalon, Bitfury,

Bitmaintech  all do a lot of bluffing as to where they are at.

Bitfury showed us a 14/16 size chip  nice demo  so where the fuck is it?

I don't know.

Bitmaintech says zip.

So I don't know.

Avalon said we are not making one yet. Now they say they are making one.


Here is the deal.

If any company above has them made right now and keeps very quiet about them.

They can simply replace all their older gear. Not raising the hash rate a lot.

I have 10,000 s-7's  at 4.7th = 47ph burning  13,000 kwatts

I put in 10,000 s-9  at 6th = 60ph burning      7,200 kwatts

So hashrate stays stabile  not much gain  and  spend  5800 kwatts less.

plus bargin sell my paid for old s-7's  or my paid for avalon 6's

So all three companies could have a lot of running .10 watt gear or none of the companies or one of the companies.

Being secret right now with ½ ing coming is to their advantage.

We get to watch a high stakes game between them.  Plus  while they keep hahs rate stabile we make a little coin.

I could believe that 400 or 600 ph right now is all next gen. mining as I type.

Or under 10 ph is next gen mining as I type.

I still have to mention I also got the email from Avalon and if  blockC wants me to do a group buy for them in the future I will do it.

Only time will tell but it sure is encouraging.

this time i might try to get in that group buy if you do !!! or try to  for at least 2 miners , i do like Avalon stuff much better after owning just one A4.1 wish i had bought them back then instead of S5. power is not a problem  or won't be i applied for a lower rate @ 0.26 lower per kwh then what i have now for the next 19 months which if a lower plan comes a long which is very likely in my state, i can drop this plan for it with no cancellation fees, .+ the way it looks by June, I'll be solar. They told me today the planing is over some time next week i should get a email with the final plans to be approved my me then it's a long wait on the state for permits.

What I like is this

 16nm SHA-256 mining chips for everybody!

I do believe that we can count on that, i dc if they have it now as long we get it with a lot of good use left.

On glitch in doing a group buy is that kilo17 is out of the BTC game. BlockC really screwed him over when they backed out of his Avalon6 chips deal with them. Within days of that he sold off his 300TH farm mostly hosted by BlockC and told them to get stuffed.

I wish I had known this. If it were posted I must have missed it.

WoW  :'(

I missed it to, I'm surprised he is out of the game.   Would not have seen that coming.    Do you have a link to this happening? 

Just would kinda like to see the post's as I missed them.  Thanks for posting though as I was clueless of this.

My info on it came from Sidehack a month or so ago when in a pm on the Community miner I offhandedly mentioned that Kilo had dropped off of CKpool a couple weeks earlier. Aside from that there is kilo's last post to the Forum https://bitcointalk.org/index.php?topic=1291890.msg14495583#msg14495583

For his part, I'd guess that kilo took a professional approach to that cold slap in the face and after selling off his hardware (and hopefully getting back the monies he had already given them for chips) just dropped the matter and left the BTC scene without ranting about it here and prolonging the burn.

 I haven't seen hide nor hair of him since even on Skype where he and I were talking a lot about the Avalon6 project :(

 It's been indirectly brought up in other threads regarding regarding Avlons amd BlockC by myself and sidehack as well.

That is a shame as kilo was and is a solid person.
Yet another sign that I would prefer be pointed in the exact opposite direction.


Too bad that happened I truly liked him.

Yeah, that chip deal was in the works since at least February. He wanted a certain number of dead boards for a project he and I were working on together (which morphed into the community design), and they actually promised to send him about twice as many as he wanted. Which was not a bad thing. And apparently kept dragging their feet for weeks and then bailed entirely. If anyone's wondering why I didn't have progress on an Avalon-based system weeks ago, that's a pretty substantial contributing factor.

That is a true statement- but I am only out temporarily while waiting on the chips to come out  ;)

Just haven't had a lot of news to update really and I don't want to post just to post.  On a side note I am in contact with the Bitfury ASIC supplier weekly and as of last Monday chips were still a no go

thanks for update kilo.

 Sad about the avalon issues.

It has become really hard with so few asic chips available to get any one to sell them.

Hope to see the next gen miner before end of july. Bitfury, Bitmain or Avalon ?? Cant wait to see.  ;D

Really glad to hear from you again! Also glad that mining is still in the cards for you (sans Avalon).

Very glad you are still pursuing the ephemeral BitFury chips. Also for reasons I've previously expressed about TSMC and 16nm chip production, not surprised on the lack of movement. But - we should be getting close.

I do hope so.  Would be nice to chips and a custom build.

It is good to hear a Update and original story to.  Thanks for posting.   I think getting chips is going to be hard no matter what at this point.  Unless someone gets them like LKETC that has a history of unloading tons of chips to others, I think most keep pretty internal.  For better or worse.
 
Still wish you best of luck though as a community miner sounds very cool idea.   But I imagine it's quite the uphill battle to get chips in current market.

Sure, but Avalon have an actual history of delivering their products to consumers. While Bitfury has a history of taking pre-orders, only to refund you and build their own farms instead.

They did that to me and quite a few others.

Basically the asic builders have a lot of leverage over the game.

That's actually an interesting concept - we actually don't know what control they have over their own company. Consider BTCS' string pulling over Spondoolies and that we only know about it in detail because BTCS is a public company. Bitfury similarly has significant VC investment and its likely that that investment comes with control.

We don't know if the V's actually get to decide to release a generation or not and its not unreasonable to consider that the investors may usurp giving away their very expensive chip for little.

Good point for all we know bitmaintech could be 100% controlled  by Chinese government or investors.


So while it looks like asic builders are doing their thing so to speak it is not really the case.

Well... I do and that's not as true. They raised they money day 1 and it was a reasonable amount of money, and then make the majority of their 'resources' from trading.

but don't they need permission to do that stuff in china  i hear it's bitch to tour there all the bs they put you thu for a visa etc and then you just can't go any place you want to, so i wouldn't put past them being controlled by gov .Ive often wonder how much the gov is in control

On the other side of the coin.

I grew up in New York City.

I knew of people that owned entire City blocks of Queens and Brooklyn all via a handshake.

 All government  documents did not mention the handshake owners.


So the real power for asic  can be in spots no one would really know,unless of course they are the actual holder of the real power.

If Putin sent someone to talk things over with bitfury and wanted an off the record deal with them it would happen.
If Bush/Cheney wanted a off the record deal in the USA with some company it would happen.
If the leader of China anted an off the record deal in China with some company it would happen.


Ghost ops flow in both  directions .  So the real power is hidden.

But on the surface  our game looks like asic builders.






  So I would prefer someone figuring out how to program a gpu to whale at silly good numbers mining BTC  as I rather use gear from real companies.

I don't like alt coins but I see what eth coin does  and I can see that  a good gpu coin could always exist. 

The eth coin market cap is about 700 million  the btc market cap is about 6300 million.   

 Does not take much to see a fundamental issue for asic builders vs gpu builders.

Asic builders attitude to buyers is pretty much fuck you show me the money.

Gpu builders  give service and warranties that are real.  Just a little tip and btc could slide while eth coin jumps.

The security issue is pretty much equal.

   We never had an alt coin go to numbers like   .1 btc per coin  the ½ ing could do this or not.

I  Will be watching and waiting for gear from avalon.  I also tell them sell us chips  believe it or not it is a safer way to earn money for your company.

It doesn't take much to get a visa, and even then you can for the most part go wherever you'd like within reason as long as you adhere to the visa terms, basically the length of stay and number of entries into the country.

To get a visa you usually just need a reason to go, having a ticket or a tour is usually a good enough reason for one to be issued, but there are companies that you can pay and they will arrange all the paperwork for other types of visa's with less restrictions (ie more entries)

Let me tell you, its far, far, far more difficult to get a US visa (for anything other than tourism) than it is for China.

Oh yeah it's much harder if you need to get a US visa as in to come to the US. It only took me a week and a half to get a business visa to go to China and about $400.

I ended up not needing one and did not go there.  But when looking China had one that offered 10 years.  So you can also get a pretty good length of time if you had proper reason.  I was surprised personal they offered one for such a long period of time.

I think most are hodling and waiting at this point for nextgen. Although I am also sure some people are picking up the odd good deal S7 and a6. Personally I and others have mentioned the stale air accumulating around the control board in the A6, but that is easily addressed.  

I had fantastic support from Spondoolies and wish they were building something competitive for a home miner. 110 TH is too many eggs in one basket and I do not see them giving us anything smaller with BTCS in need of as many which can be built and shipped. This is pure conjecture on my part except for the above average support, but it isn't hard to be above average when the competition was Bitmain.

Having said that I do see effort being made to use your regional support, I have purchased miners through the Bitmain regional guys in Colorado.
The last time I checked pricing it was extremely nice to purchase a new miner through Yoshi's group, get fast support communication, but... There was still a delay in actually having the miner shipped along with having a repaired unit shipped. Actual shipping time was fantastic, but even though the unit was very obviously here in the US there was still some hangup between my BTC being sent within a few hours, and the unit shipping.

...and Having said that, Yoshi stood behind good customer service, apologized for the delays, and I will not go in detail, but he took care of me.

However, I am interested in a bulk purchase and my overall experiences since the first miner I purchased from Bitmain, the S4s hurt like hell because of Bitmain customer service, the batch 1 S7s, even though they gave compensation the people who received theirs before the Holiday had a significant advantage. I had a very bad taste in my mouth after spending 9+ BTC for my first S5+, the fan die within a week and the factory telling me SOP is to buy your own fan locally. Other Bitmain employees told me I may as well do the same thing. Again Yoshi took care of me. I bought a spare at that time as well.
Yoshi has been the consistent reason I went back, and when I make my first real bulk purchase, it will not come down to who hits the market with the most efficient gear first, it will be the choices available to me based on past experiences.

I still haven't wrapped my head around why Bitmain has not set up a true factory leg in the US including assembly. What a jump it would give them to have a US team able to support and sell here. I understand they would lose some profit margin due to labor, but I firmly believe that it would be made back quickly with volume AND they could offer next day air service for people who want to pay it. Bulk orders could go out Fedex freight or an equivalent.
I guarantee they would sell many more spare parts. Look at how much shipping is added for each additional part from the website now. That has been the sole reason I have not already picked up several hundred dollars worth of spares.

They could also sell hash boards. Simply have larger runs made. Look at the people who would pick up multiple boards so they could build their own unique designs. People like me would purchase a hash board even if I had a warranty board going back in to be replaced to reduce downtime.

As was mentioned if BlockC approaches these things, lowers the MOQ to 1 for the small miner, and allows each individual part to be purchased I see a real fight starting.

The same potentials exist for other companies of they stop talking vaporware and actually make a move.

Of course I am sure there are many items I am not seeing, but I cannot understand why making your brand is not the long term goal.

Sloopy,
Thanks for your post, very well written and makes perfect sense.

For the same reason Spondoolies couldn't compete on price, its just too expensive. If they wanted a US facility they'd continue what they had before and make them in CN then ship them over in bulk.

They already enjoy 2 day air shipping to most of the world and would need a farm in the US (expensive), so without a significant bump in price to pay for these niceties its never going to happen.

Thanks. I appreciate the kind words.

Re-reading it in the light of day there was one main point I wanted to make in this thread which I did not.

Is it just me? (and it wouldn't be the first time), but seriously is it just me who looks back on the way the Avalon 6 versus S7 pricing structure and MOQ were lined up seemed to coexist well?
Similar to how AT&T and COX / Time Warner may coexist and both do extremely well?

Maybe I missed it, but I purchased an A6 much later than others and paid well over what the S7 cost. I did this knowingly. The same way I pay my bill for Internet and Cable TV every month. (Never watch the cable TV)

I wonder who gets the credit for those things aligning so perfectly?
Bitmain maintained the lower price. Many times to the point it was not a good decision to do anything else. *IF* you had reasonable confidence the hash boards or the entire unit wouldn't need to go back. So a gamble on top of your BTC mining / price gamble, just add that layer on top.

I do not mean to snub either company. Good Business decisions. Actually, so smooth it is almost a concerted effort to maximize profit and keep the price at $750 - $1100 for a long time between November / December and March / April.



Regarding making them in CN and shipping them in bulk they are still doing that today, they just do it through Yoshi. Very little difference. I am talking about a true extension where there is true assembly. That seems to be the only way to have every single part available in a next day air fashion.
Does anyone have any numbers showing where what percentage of miners are sold to which countries? Obviously China is #1.

These "niceties" are what make a huge difference in many people's purchasing decision come true competition, and why so many people support the community miner project. These "niceties" are what will separate people who have a job and people replaced with automation or they have no market share in the decades of Industry to come. China is growing in the middle class, and they will demand higher wages, there will be a change there. It may be 20 years, but it has already started.
The simple fact is you get what you pay for and I can show plenty of people who have parts made all over the world but assemble and test the product in the US (or their home country). Are there shitty companies in the US? Sure. Does it take weeks to get a replacement part? Not if they want to sell another item and shitty companies do not stay in business unless they are monopolies or government funded.

They may "enjoy" two day shipping of new units to some places, but who has been getting their unit repaired, who has been getting a fan replaced in a reasonable time? Who has received a hash board in two days? HUGE difference.
Do you understand now dog? If not, just forget it and move on trying to bait someone else or get back to loansharking.

And as mining history has shown us, people aren't willing to pay a price premium for frills when there is raw hashpower available elsewhere cheaper.

Can you have an adult discussion for once, please.

???

Except for everything Spondoolies ever sold.

The SP10 won by efficiency and density, not by frills. The SP3Xs won by preorders, not by frills. The SP20s 'won' by being sold at a huge loss, while also not being very good.

So you wouldn't consider the slick rackmountability, total adjustability, integrated server PSUs, spiffy UI and everything else Spondoolies did (and charged a premium for, then sold truckloads of) that Bitmain doesn't to be "frills"? Please, define "frills".

$/GH/s and W/GH/s matters to Bitcoin Miners. The fact that Bitmain is thriving and Spondoolies-tech died is confirmation.

Spondoolies-tech died because of mismanagement, not because their miners were to costly.  Selling the company to BTCS is and will be their demise.

Well it was part in cost, as they tried to compete with bitmain and claim they sold them at a loss.  Which obviously does not work good.    I think it is true they just could not compete.

Now you could argue management to.  But price was a factor for sure to.

I would not consider it successful, no. The buyer has to pay for those frills, and unfortunately Spondoolies wasn't able to recoup those costs and still be able to sell units. As scyth3 suggested, no frills won over frills.

Bitmaintech won over sp-tech because of power cost not gear cost.

If I build and mine at 2 to 5 cent power cost and you build and mine at 9 cent power cost you will be ground down over time and lose.

But wtf do,I know I know this at 1 cent power an s-3 still turns a profit.

Sp-tech was never able to self mine the way bitmaintech did.

In the long run they died off because of this more then any other reason.

I could make Avalon win over bitmaintech if I supply them with all the free power they want.

It is obvious bitmaintech has a huge source of cheap power for self mining which allows them to sell gear cheaper then anyone.  The current price of 433 for an s-7 after coupon is a fine example of how they leverage the low cost power to rule the world of Asian asic mining.

Avalon is content to be a distant second since it works for them.

I would not mind Avalon picking up their game a bit.

But I rather see some chips sales more then any other thing.

Spondoolies made a conscious decision not to self mine, even though they ended up building a sizeable farm. That's nothing to do with power cost, and if they had decided to self mine they could have created any number of global mines in the same locations Bitmain had access to.

So, Spondoolies lost because they decided to not openly screw over customers (from both a design/feature and service standpoint) in as many ways as possible and still retain customers? Sounds about right.

yeah  no matter how you phrase the language  Sp-Tech mining it sp20 at 3-5 cent power would have given them a real cushion in the game.

They were not able or willing to do this  which is why they lost out more then any other reason.


If I build my gear in Howell NJ   and drive it to my 40 acre solar farm in New Egypt, NJ
a short 7 miles from my house I have a huge edge over other builders.
(note this farm is not yet built or is even soon to be built it is merely possible to be built)


If I stay stealth   until my farm's capacity is maxed out> I can then sell my gear at the lowest price on the market.

Basically this is what bitmaintech does.
To further their edge they do not sell chips so that a guy like sidehack can build a  miner that is versatile.  Say .19 to .31 watts a gh


So far bitmaintech has pulled this off and made  $$$ .

Should be interesting to see what the ½ ing does to this.

 :)

Interesting thread. Any questions?

When will we see a 16nm Avalon miner? What is the target $/GH and W/GH?

When will it be available?

It may have already been asked (that was sarcasm, by the way)... when are we going to see a working new miner that we can purchase? :)

Is there any consideration for upgrade path for existing mining hardware - i.e. board replacement?

I have 8 avalon 6's

It would be nice if the board could be set to the old heatsink and the case, fan, internal controller all were used along with the 2 heat sinks.

The savings on shipping would be huge.

Agreed, Phil.  That's why I asked :).

Yeah upgrade board could be awesome. I do think that home mining need to go in that way. Reuse old heatsink and braket.

Impossible.

A7 is a very different (much much more optimized ) design. In both ASIC and system side. So... like every time...  :)


Sooner or later.  :)

Any way I can get any information from you guys about the A7 chip? Or even the A6 chip? Not a single one of my inquiries about anything have been responded to since last June so I gave up after about six months of trying.

I had set up a big sale of Avalon4 units between Canaan-Creative and a guy I was consulting for who was starting a farm. He ended up never paying for the order (a problem the guy had often, which is why I use the past tense in regard to my dealing with him; I was tired of facilitating his laziness screwing people) so I assume I've been officially spurned on my deadbeat client's behalf but I'd really like that to change.

Sense miner design and production become a very difficult (mainly about cost, timing, supply chain, manufacture, software, etc...)job, I hardly believe any 3rd party can do a competitive design. I think anyone (who is not in present industry )should give up for self-miner-production-for-self-mining . Or you have different opinion?

If you sold him a7 chips with documentation he could build good gear.

One problem with the a6 vs the a4.1 was it was locked via software.

That works for many farms but for some of us an unlocked miner has more value.
The problem with just about all chip builders is they rather gear to bigger farms.

I did a lot of work to help sell a6s to the guy with limits to cheap power.

I am in a joint venture on a larger but not huge solar array.

The power is cheap actually cheaper then most can do.
If we mine 8-11kwatts  year round the net power cost would be under 2 cents.

But that is all the power we have on the 2.5 acre spot.
So an a7 with the downclock under lot option is better for us.

We could just buy a few more and down clock while the starting cost would be higher for us in the long run the better efficiency allows for more hash power.

10 at 10 th using 10 kwatts is 100 th
12 at 9 th using    10 kwatt is 108 th
Both bitmaintech and Avalon did not offer easy under clock down volt the last round.

I know side hack could build a more efficient miner with quality chips.

A lot of people are anti Asian when it comes to asics.

I personally do not care where gear is made or where chips are made.
But I do care that you won't sell chips to people that want them.
By you I mean every ASIC builder simply won't part with chips.

I understand you want to make money , but having your chips in other peoples designs would make you money not cost you money.

I also understand you can't tell us all the why of your selling tactics. There are restrictions. Set by your company I don't think help your company long term.

Just remember at least 75 percent of all ASIC gear is in China.
Not safe to have all your eggs in one basket.
Suppose USA elects Donald Trump and he decides to default on the debt.
We all know,he has done this time and time again.
With his businesses.

Since USA owes more to,China then any country it would rock your world.
It would be nice to know more of your gear was in another spot then China.
In case there is a lot turmoil due to the default.

Also multiple plants with chips and build gear protects you from earthquakes floods etc.

I like building things under 500W. None of the major manufacturers have released a 500W miner in the last year or so. One of the design ideas I've had for over a year, if I had 1000 boards right now with even the same efficiency as S7 or Avalon6 I'd be shipping them all out tomorrow.

By only building miners in the 1KW+ range, mining manufacturers are alienating a lot of the customer base. They're focusing on the wealthy (or gambling) few and ignoring everyone else. From a pure profit standpoint it makes sense, since one lump sum is always more desirable than a thousand small payments; having one customer to deal with is better than a thousand customer; the guy willing to put in a million dollars will probably have more intelligent questions than the average consumer. But I'm an average consumer and I generally give very few rats-asses about what the wealthy are doing so if I'm going to build a miner, I'd just as soon build it for the guys like me.

That said, I am probably better equipped to build a miner than most, given six years at University with three undergrad degrees in computer and electrical engineering, plus I've spent the last two years building up a small electronics manufacturing facility. I agree, miner design and production are difficult; I know this because I've been doing it. The thing about my "competitive designs" are, they're not competing directly with what the big guys are producing. I'm aiming for entirely different market sectors that are, at present, left completely bare of options.
I'm not going into miner production for self-mining. I could, if I wanted to; I could clear all the customers out of my 100KW hosting and stock it with my own gear and probably make more money in the long run. But why should I, when I can sell the miners, make a satisfactory amount of money and help other folks stay in the game when nobody else seems to think they're worth the time of day?

At the end of the day, selling chips to guys like me still makes you more money because you're not going to sell fewer of your 1000W machines. You're going to be getting paid for the chips that go into the 100W or 400W machines we'd be selling to the people who wouldn't be buying your 1000W machines anyway.

So far, Bitmain's the only outfit that's actually gotten me chips. That was over a year ago. They were unwilling to even talk about BM1385, even after the S7 was released. To be honest, I'd rather not seek chips from Bitmain as I disagree with some of their business practices. They don't start selling anything until it's good and ready to ship, or at most about two weeks out. I respect that. But price gouging, profiteering, rampant self-mining, unreliable designs and selling obviously used gear as new I'm not so fond of. That pretty much leaves Bitfury and Avalon. Up until the A3218 generation, Avalon had a history of working well with independent developers and building pretty solid hardware so I'd rather work with them if I had a choice.

That should be an adequate summary of my different opinion.

To be honest, very hard. In history, few of 3rd party do a competitive job. 16nm Chips are extremely hard to deal with, also.


So what's your need and what's your want?
All you want is a downclock/downvoltageble miner? If so, it's possible.
If you need a miner under 500W? It's possible.
any others? :)


Miner develop needs a large number of technical support by us, we do this in avalon 1,2,3,and 4. NO successful cases, except for some clones in first and second gen.


This is business competition, Chinese manufactures just win this round. For miner design and production, I don't see there is any chance for out China vender (now). Also, China is really really big country, don't care about earthquakes and floods. :)

This is getting interesting -> Spondoolies out Avalon back in  ::)

Make a miner that is better and cheaper than Bimain's 16nm would be my request  ;D

They were never out ;)

There have been no decent miners under 500W released since the Antminer S3 in summer 2014. I'm not including Technobit's Minion boards in these considerations for the well-documented reason that that offering was mostly a scam; in any case, it was also summer 2014.

There have been two consumer/non-rackmount miners released since the S3 came out that allowed downvolting - the Avalon4 (fall 2014) and SP20 (winter 2014).

Nobody said it wasn't possible to do this. What we said was, nobody's done it in a year and a half and it's about friggin' time somebody did because it's what a lot of us want. What I'm saying is, if you don't want to, sell chips to someone who does. You make money from the sale of chips; with the exception of providing technical information to a very few technically competent people there is no customer service required on your part. If those chips go into miners designed for a customer base you are not serving, direct competition is minimized and you make money off miners you didn't have to build, you don't have to support and that are going to people who wouldn't have been your customers anyway. It's probably worth thinking about at least.

This is your mistake. It isn't about money. It is about respect, and by "respected" it one from the Occident should probably think "feared". My shortest explanation is that ngzhang belongs to a class of people for whom having 90% of $1000 is more than having 10% of $100000. It isn't about getting more money, it is about getting more respect. It is a fundamental difference in motivation.

He can't do this because it is too risky for him. It is possible that your design would be better than his. That would mean a "loss of face" or loss of respect (and self-respect) for him. That's why he can't and wont risk doing what you propose.

Well put My brother has often tried to explain 'respect' and loss of it in the Chinese culture to me.

My brother has spent 5 years in Taiwan and 3 years in mainland China.  He was a dean at a NYC college he speaks reads and writes Chinese and has a PHD from Columbia in Linguistics .  Pretty much really smart in comparison to me.


The chip  in a design other then Avalon 's own design most likely will not happen  for a lot of reasons including your respect one.

As I see this from the viewpoint of I don't need to mine for money.  I don't try to make a lot of money and I really want to promote small mining as much as I can.
This is just not an agenda for any asic builder as the small miner with 'free' power kills big farms.

Many people can do a small miner at their office legally. By small 500 to 700 watts not noisy for space heating.


Many homes in the USA have electrical heat old school plain and simple radiator type.
In my area  there are two senior only developments they use electrical heat. four room small homes with 1 bath.  They would be ideal for 500 watt miners like an s-3.
Quiet and they just  could tie into a network in the club house of the development.

Both cases above put a hurt on large farms. 
As for sidehack building a better miner then avalon would build he probably would  in adjustability  and efficiency .

Frankly I don't sit in on a 25ph farm and listen to them discuss what they want with Avalon. They work from a different viewpoint. Then I do.
They would make the coin and sell the coin and leave the system in the dust.

But the crypto world will adjust  look at  Eth coin   it is an adjustment to  BTC asic builders.   The gpu companies have gaming and design as 1 and 2 but they have come to realize mining can help their bottom line.  Eth coin will switch to POS is Eth asics come out.

then a new coin with a new algorithm will take over. always keeping gpu's as top miner.
In time this could simply kill off asic btc.   Why because they won't sell chips to different designers.

Many companies make gpus and the major chips in them are more accessible to designers.  so by hoarding asic chips  the major players are risking the future of btc.

I spend 1400 watts for 150 mh to mine eth coin  even at my 17 cent summer power rate I turn a profit.  plus have good pc's to sell if I choose.
Plus I know that if an asic builder builds an asic chip the coin goes to POS.

Mind you I will still buy an avalon 7 and an bitmaintech s-9 for the array.

But my agenda is small miners in house as it keeps the gear everywhere. So if rotating from BTC asics to Eth coin gpus allows miners to be more wide spread I will push Eth coin.

As an aside  AMD  has more then enough money to prop up the price of Eth coin buying it on exchanges .  It will help thier gpu sales.

Completely off topic, but this is one of the reasons I don't support any altcoin at all. The only coin worth supporting is one with long-term viability. Mining a coin you expect to die and be overtaken by another is working for short-term profit and in that scenario, when you profit someone is left holding the bag.

Anyway.

If I built a miner better than what Avalon themselves could produce, it's only because I have a different definition of "better" than they're operating with - a different set of priorities. Based on everything from S5 to Avalon6, it looks like the big manufacturers are now prioritizing a minimal parts count and maximizing efficiency. That makes sense, to a certain extent. However, when you run your ASICs at the top end of their efficiency curve you leave a lot of room for improvement (at the expense of more ASICs per hashrate; see S2 1J/GH 1.57GH/ASIC vs S1 2J/GH 2.8GH/ASIC with the same chip). If a miner had a stock setpoint of the chips' top-end capability but allowed for adjusting the clock and voltage, the efficiency could be improved by 30-50% (depends heavily on the chip of course) at a cost of hashrate. But if the miner is priced to break even at top clock, after that any extension of its viable life because of undervolting means more profit. As recently noted in another thread, the year-and-a-half-old SP20 is still viable for some consumers because its adjustable efficiency range is from about 1J/GH down to something like 0.45J/GH, which extended its life substantially and kept it viable through the next two generations of competitor hardware.

In the Avalon4/SP20/S5 generation of gear, I didn't have available funds to buy miners but of those three the only one I actively wanted was the Avalon4. Because it was quiet and allowed for voltage adjustment. Both of those attributes are now considered unneccessary features but if you took a poll of users here you'd likely see an overwhelming response in favor of machines that run quietly and have adjustable voltage. If you want respect from the community, the way to do it is not to ignore them, but to facilitate getting them what they want and doing so graciously. When I got chips from Bitmain there was quite a bit of positive PR in the community for them being willing to make that deal. When Bitfury announced they'd be selling their new chips to third parties, that was the talk for the next few months; when in the previous generation Bitfury kept them all to themselves that was grumbled at for most of a year. Just saying, whether you want money, respect or both, supporting the development of a miner within the community and geared toward the small miners would have an incredible return compared to the amount of effort required.

At minimum , I'd think that potentially , "licensing" production onward to smaller regional players could make sense. I understand the sentiment, we've seen so many bad entrants, failed projects, defunct companies come and go. However, in earnest, not since BitFury 55nm has anyone actually stepped up and supplied chips to the people that want them. ( At worst,  you sell more chips , others fail , CC proves correct )

We love our A6s , it's my favorite current gen miner, how soon is soon for the A7 in your mind?

Well, I think there are many viewpoints and opinions, I suggest please email love@canaan.io with your ideas and business proposals. Canaan is open minded. Please follow us at https://twitter.com/canaanio

 :)

There is a very long thread here about myself and SideHack producing a miner using Avalon chips- it just so happens I got screwed on the deal by your partners at BlockC and was promised chips then they decided I could not have them.

So the deal is still on the table if you would like to help us.

Keith

I think there is probably a larger demand than most people realize for a real alternative to bitmain.  I would even pay more (as long as the price still made it profitable in the end) to support an alternative manufacturer.

I have purchased about 8 or 9 avalon6s for myself.

I helped sell about 40 in group buys.

I like the company.

I would love them to set sidehack up to build smaller miners here in USA .

A 300 to 600 watt model.

I have invested in sidehack's compac sticks.

I would invest in this idea.

I would also help to distribute a smaller miner if it is ever available.

Im like Phil, I believe in Sidehack and already invested in his project.

I bought avalon 4 and 6 and did some review. I also helped for the distribution of Compac in Canada and will be more than happy to help the distribution of the next gear.

Come on Avalon, give us some chip to keep the home mining alive  ;D

for these reasons .. they will not send sidehack chips.> It cuts into their bottom line.. Instead they would love to sell you a miner ;) ;)


i bet they would send them to me :D :D  They prolly think id get stuck with them and wouldnt hurt their bottom line.. You guys airing this all out in the open has been whats hurt so many projects in this space to date.. Maybe now you will think of this before hand.. They see that sidehack will most likely be able to pull this off thus fucking him on his chips.

EDIT> also since its the sales of miners that is keeping this afloat.. Sidehack would have to be in agreement with the other asic manufacturers on how much of sales goes back into propping up the price to further future sales

Best Regards
d57heinz

Well you may see it that their bottom line gets cut if they sell 10000 chips to sidehack.  They may see it that way.

Frankly  they shut the door on the guy that wants a decent lottery miner.



How about  a smaller unit for the guy that can run 400 watts all year long for free.  As he would be using a space heater anyway.

Why shut that guy out?  No need to do that. 

  let sidehack build one here from chips

and copy it from sidehack to sell it in china.

Both China gets it and USA  + Europe gets it.

Like I've said for a long time, I don't want to compete with 1KW units. My target customer is the guy who wouldn't be buying their hot loud machines anyway, so they'd be getting a cut of sales they wouldn't have had otherwise.


If by "copy" you mean "license" or some other formal agreement for IP sharing, sure. If by "copy" you mean "steal", to hell with that.

I would mean you do an agreement that in turn for chips sold to you and support on chip  pin out to pcb layout.

once you make the design they could use it in China/Asia

 you could use it in USA/Canada.

Small ½ power models say 400-600 watts.

We all know they do bigger units but if they want to supply small ½ power units to China /Asia it is a natural to share the design with them.

While you handle USA Canada South + Central America.

You did a deal with a europe guy for the compac stick    so something on those lines.

You agree to only do 400-600 watt machine so as not to compete against their 900-1200 watt machine.
You don't sell your 400-600 watt machine to China /Asia you license them to do that.
They only sell they 400-600 copy of your machine to  China /Asia .

It is a simple easy business plan.

I'd be up for an agreement. I won't design a consumer miner with a stock setpoint above about 500W DC consumption, so yeah competing with 900+W is by no means what I want to do.

To be transparent, I did license the Compac design to MacEntyre for $2 per unit of whatever he manufactured, in addition to the understanding I'd send any Euro customers his way.