Topic: Antminer S8 ? (Read 13585 times)

No this is Antminer S7+ 12,4Th/s

 SP20 was more efficient than the S5 was when hashing at the S5's hashrate - though it was VERY close, close enough that manufacturing tolerances probably had a few of the worse efficiency SP20s not managing that against a few of the most efficient S5s.

 Bitmain will probably already HAVE competition from Bitfury on efficiency, and that LK1402 might be as well depending on how close BW.com/LKetc hit the announced specs on it.

 Right now, though, it's still looking like Bitfury is in the lead, as they've SHOWN working silicon for the full-custom 14/16nm generation and nobody else has (I discount KNC's "Solar" announcements, they've not shown any indication they have fielded that FOR REAL in any quantity worth mention IF AT ALL and KNC has a long history of paper announcements with VERY VERY long delays before delivering actual product).

(edit)
 And KNC is bankrupt now and likely to soon be dead anyway, RIP no big loss....


 I doubt that the S9 will be 1 BTC/TH - though I won't be shocked at .6-.8BTC/TH range. Consider that the S7 is down to around .3/TH lately - the S9 would have to be a .10w/GH unit AT THE MINER (not just at the chip) to be worth over *3 TIMES* the price/TH.

which would make it pointless, and too costly none will buy it, unless he want to sell to a greater fool than him

10 tera can give you only 0.025-0.03 so you see roi will be impossible

Most efficient publicly sold gear, we've been hearing about more efficient chips than theirs for a while now, but with no chances to buy any.

Price easily will be ~10BTC for 10Th

The s-9 will not see much competition   in terms of efficiency .  Basically their gear has pretty much always been the most efficient.


Now the price point  most likely too high.  But people will flock to this since they know the ½ ing is coming.  I will order 1 just because.

Good spot, thanks for posting this.  So the only question will be now... invest in the S9 or will BW release something better (14nm) shortly after that will be even better than the S9?

Thoughts / speculation?

I missed it at first look.  Thanks for sharing that is interesting information for sure.  An official response I was not expecting.

https://enforum.bitmain.com/bbs/topics/3433?page=1


^^^second from the last post has what I would assume is confirmation of the S9 being released very soon.

It's a cloud site trying to get buzz with this sale.  They offer NO details on S9 let alone a date it comes online.  So they are offering a certain amount of a S9... they don't even need to know final specs to do this.

It all seems very shady to me personally.  I would not pre-order "Mining Shares" just due to them putting up a S9 sale that offers no details.

can you buy it directly? what are the spec of this new s9?

Rumors on the BTC ether:

S9 builds already out of stock, the Swedes already bought them all.

https://www.miningsweden.se/product/mining-shares-200-gh-s9-antminers/

They forgot some zero's here & there in the descriptions of the hashing power to rent via the so called S9 capabilities.

Mining Shares 200 GH – S9 Antminers

$49.95
SKU: MS9200 Category: Mining Shares

Company-share
Mining Sweden Antminer S9 2000GH Shares (MS9)

$479.00
SKU: MS92000 Category: Mining Shares
--------------------------
Product Description

We have now information about that the new S9 miner soon will come out on the market so now you can preorder your mining shares with the new S9 Miners.

2000 GH S9 Mining including all costs for 12 months.

This product doesn’t start mining before we have them in the farm

I agree we are not seeing them hit their best efficiency because the main market for them has very low electricity cost and what matters is how fast they hash rather that the last ounce of efficiency. S1, S3 & S5 all were able to be pushed to significant efficiency improvements and I am sure that S7 could see a big gain as well, but they would rather just sell the next generation.


Rich

In theory, if you set up a miner with a buck setup to run a string of the new BitFury chips, it COULD hit the mystical 100w/TH mark - but I suspect most miners using it will opt for fewer chips = higher hashrate per chip = lower $/hash at the cost of efficiency.

 Same reason you don't see the BM1585 (or any other mining chip Bitmain made) getting close to IT'S max efficiency - fewer chips = lower cost per TH = bigger performance = more folks WANT the miner even if it's not max on efficiency (which is what killed Spondoolies, they made GOOD tuneable miners but too many folks wanted low cost/MH).

I still think we're a year off before we reach effective 1TH/100w - at least on 110V miners. 240V rigs will be closer to that goal.

i'm expecting 10 TH/S for 1000 watts   maybe around nov this year after halving

June at least, or earlier

 Yeah, I've had one PS die already (no idea how old the miner was, used unit from Zoomhash PS died on first turning - they DID compensate me for it, bit of miscommunication issues but no real hassle).
 I find a Seasonic X1250 is a VERY VERY good fit in place of the stock unit - it's a bit longer, but the fan lines up with the hole in the side of the case well and the modular connectors you NEED to put in have enough clearance to work well.
 EVGA 1300 G2 does NOT fit well (it's a little longer than the X1250), the fan hole doesn't line up nearly as well *AND* clearances on the cables are so bloody tight you pretty much would have to move the controller board to avoid shorting stuff out from cables rubbing against sharp things pretty quickly.

 Both have PLENTY of power to run A2 Mega Terminator units as a nicely high overclock.



 I don't think an S9 by July is likely - but I wouldn't mind being proven wrong, and it's certainly not impossible (Bitmain DID announce they were working on their next-gen chip at the same time as they released the S7 after all).

all the release dates, tech,etc etc is all going to depend on BTC price after halving.

what cat these are simply preorders on cloud mining.

100 share of 20gh = 2000gh or 2th  of a s-9  that does not pay until they come.


 2000gh for a year earns about  705 back

 without a 1/2ing


they want 598 for the 2000gh good luck with buying that.

Only proof right now is on the Mining Sweden website.

They did S7 hashpower preorders a month or two before the S7 got released. So if you go by that, then it could mean the Antminer S9 could be on sale within a month or two.

Difference is this time Bitmain are keeping quiet on it so whether Mining Sweden have just left the cat out the bag bigtime, who knows for sure.

Not rumours.

Mining Sweden are selling S9 hashpower preorders on their website. Make of that what you will.

I looked in my crystal ball and saw the Antminer S9 on sale before July

 

There are rumors that some Sweden got S9. I was not told exact name of pool - but Sweden was pronounced

Maybe that is just a campaign to drop S7 price down

I never heard that antminer release a s8 and i think it needs more power consumption. and i think its not recommend for now.. Because its not profitable unless the s8 is cheap and the your power is free..
S9 impossible never heard it..

proof ?

Apparently MiningSweden has bought S9s
the launch should be just around the corner

Wouldn't be an S8 - and I doubt there ever will be an S8 for the same reasons there never was an S6.

 I could see a faint possibility of an S7- :

 More chips per board, running the BM1385 at it's most efficient point.
 Probably 2/3ds as many strings as an S7 due to needing more chips for the efficient string operating point.

 Very long shot, 2 boards instead of 3 - but that might be an "S7- Lite" instead, which would be a LONG OVERDUE actual home miner but the price point would have to be too low to make it viable to Bitmain as a product given how cheap they've had to drop the S7 to.



 S9 would be the "next gen" Bitmain chip - and it better show up a lot sooner than a year from now or Bitfury among a couple others aren't going to leave them any market TO share, unless that "next gen" Bitmain chip and mners based on it are noticeably more efficient than anything then on the market.


 Clock rate has almost ZERO effect on efficiency, efficiency is almost ALL about the voltage you run the chip at - thus the need for "more chips per string" to maximise the efficiency of the BM1385 in a potential future miner.

I think so, what I do not know is where you can get to with the Hash Rate & Efficiency?

This thread details a repair / mod to the S7 Hash Board that freplaces the PIC with a potentiometer that enables you to adjust the core voltage which gives the same result as changing the code in the PIC

https://bitcointalksearch.org/topic/cheap-and-simple-repair-of-s7-hash-board-1420909

Rich

So it could be done easilly?

But if they wanted to do that all they need to do is to turn down the voltage & frequency on a 135 chip S7. It just needs a change to a Byte in the code running on the PIC that controls the voltage.


Rich

The only option to make an S8 miner is to make a more efficient miner with the same BM1385 28nm chip. Now i can see the logic to create something like S8 (only speculating). After the halving, it is very possible to be hard to find a profitable miner, and Bitmain could decide to create a new miner running with a low frequency; perphaps 250Mhs or 350Mhs and a longer string. They could reach to something like 0.15 or 0.20 w/gh (maybe 0.18 w/gh)? a miner with 350 chips, doing 6 Th/s and draining 1100w?

The Avalon 4.1 beat the S3 on noise by FAR.  If your looking for quiet of that generation I would go with it, no other modern miner has come close straight from factory.    Others you would have to mod and still not sure you would get same level.

But here is my old review with video - https://bitcointalksearch.org/topic/review-avalon-41-miner-summer-mining-notlist3d-1008726 .  I was asked many times if I did something to make it quiet for video and I did not I did have to get that close for mic to pick it up even.

The S3 is my favorite miner. Very silent and strong.
I am surprised there are no S8 already. I was expecting to buy one last march but bitmain lab is silent. Probably they are testing them again for a few months and then sell it.

And they would have kept going but in a few cases I reset them due to hardware errors showing up wrong.  Run any bitmain miner for around a month and you will see a negative number on hardware errors.... which is impossible.  I normally reset so I could see real errors.   

Dragon A2's were beasts I'm not knocking them I know some data centers they were put on a ton of stress... and lived.  They were really good miners, only thing they lacked was a nice PSU.  Assuming LKETC a2 eventually a PSU dying was kinda common.

 I'd be impressed if I hadn't had 4 of my 5 A2 units run for 4 months straight - befrore I reset my home to "summer mode" and had to shut those 4 down to move them.
 The 5'th one is at 4.5 months and still going....

It was a tough sale on S5+ I think. They had much more leaks of info back then and it was known they were using it internally (and might have been selling in China but not abroad cannot remember 100 percent).   But there was enough push they did sell it to all eventually.

It was a great miner but expensive, and also a lot assumed S7 would be soon after... so again hard sale.   Power was not a big issue due to a decent amount of "hobby" miners grew to 240 and decent amount of amps.   It was definitely to big for home miner with 120 volt.

So S7+ would be good miner..... but likely high and a lot would wait for S9 seeing it as way to dump remaining chips if they do pump out S7+'s. (And could be true or not kinda a gamble just like S5+ was on that).  So unlikely we will see one.  But yes S5+ was a good and reliable miner for those who bought it.  Also held value decent as less competition when selling vs regular S5.  

Upgrade Boards are a nice idea for the end user, but a pain for a manufacturer, and typically have a higher support cost because of people opening up units and plugging cables etc. Margins are always better on a complete unit.

I am uncertain if we will see an S7+, it's an awful lot of power needed and I am not sure how well the S5+ was received? However as we know the controller board will support it and if it does surface it will be a sure sign that the S9 is immanent.

A lot will depend on how the BM1386 chip development is going? The move to 14/16nm does not seem that smooth, not helped by the Taiwan earthquake affecting TSMC and in particular the newest technologies.


Rich

My S5+ was one of my favorite miners, had a good amount of power and was reliable.  I had it run a month or more on multiple occasions.  I would be interested to hear a S7+ depending on price.   If they do it I predict same as other they use about at end to get rid of extra chips/hashing boards.

It would not surprise me if they skip S7+ though.  I think they have the design ready for S9's already.  I predict they go one fan 3 hashing board, likely were testing the design now for it.   

They've never gotten around to offering "upgrade boards" before, despite talking about doing so.
I don't see any major probability of them doing so now.


 What I CAN see happening is them offering a single batch of "S7+" about a month before they introduce their next-gen (probably called S9) miner - like they did with the S5+ shortly before they introduced the S7.


 But of course, they may decide that was a one-shot option and won't work THIS time with the halfing so close, even if they cram as many BM1385s on the board as they can to maximise the efficiency....

Likely this will not happen if they do I would guess internal.  They just have done many generations without a replacement board.  I think S3 upgrade from S1 (I believe) was first and last upgrade.

So for them not to do S3 to S5, or even S2 to S4 we will not be seeing any upgrades kits.  They are more interested in selling entire unit's at this point.

My guess would be that a "replacement board" is coming too. I mean when they will release the next generation I believe they will also offer boards for existing S7 units.
Make them S9 for example with 3 new ASICs. You could keep the controller, house and PSU (and the physical setup).

Just my 2c anyway

 Which just means Avalon is going to be VERY late to the party at 14/16nm given lead times on development and that some folks like Bitfury, KnC, and probably BitMain and Innosilicon are probably already starting design work on their full-custom 14/16nm generation of chips (and in the case of BitFury and KnC have already announced tape-out or production on said generation).



 S8 efficiency will be very close to, or the SAME, as S7 efficiency, as it will be using the same chips - IF BitMain bothers releasing an S8 model.
 I'm still betting on an S7+ instead.


 I'm guessing that the next major upgrade will be an S9 model, ballpark end of 2016 timeframe, using 14nm full custom at ballpark 0.05w/GH efficiency.
 I DON'T think Bitmain is likely to get their next generation out before the halfing, or in significant numbers if they DO manage to do so.
 Yield issues on the 14/16nm process are pretty significant to date.



 Bitmain announced a while back that there will be NO S6 Model.




 The S7 should still be profitable after the halfing - but only marginally so and dependent on VERY cheap electric to manage a profit.
 Ditto the Avalon 6, but it's going to depend a lot on if either of them handle undervolting well if at all.
 I suspect the B-Eleven will be in the same boat once it's released, for the same reasons.

 Even the SP50 isn't going to be very profitable at that point, despite it's slightly better efficiency than anything announced to date.

If they launch it in feb then yes 0.1 is ok. But i expect them to come out somewhere in april and then they have to do a lot better then 0.1.

Probably 0.05 or less.

I think it will be a 10Th+ Miner with 0.1 efficiency.

it must be lower then 0.15 since Spondoolies Tech already have a miner with 0.15 efficiency.

Could be.

I guessed bitmain left that series and will focus on the efficiency (the s5+ was a goodbye to the S6 maybe)

Too bad, since i loved the S4 and it's robust view.

After the S3, we got the S4, so after the S5 i expected the S6, but bitmain decided to go ahead with the S7 first. Since the unit is very compact i would reckon they try to maximize profit before launching a new miner.

i think that will be based on a different chipset and a worst efficiency if based on the s4, just bulkier and that's it

as i see the s8 should be double the efficiency of the s7, probably it will come out around the halving or just before it to give mienrs more margin on profit

That's the rumored S7+ as far as I know. Hopefully Bitmain does go this route.

For an S8 or S9? Maybe a year from now.


I also have an s3 stocked in here but unplugged. Power rates are just too much. That's why I too am hoping for a lighter s7.

for  now the best home use from bitmain is the S3 IMHO,and the reason why I say that is because it does not run loud compared to the s5 and s7. I know we can change the fans on the latters but still :-p
I do look forward to the new products that they will eventually come out with hehe

your ideea with a light miner seems good to me.
A miner with like 2.5 Th/s and just 700 W or so, that would be nice !

S7 is the "@home" miner, asic makers only want to sell to large company's now and they basically don't give a shit about the hobbyist.  So we have to be happy that at least we still have bitmain to buy from, otherwise we would be out of the game.  But I agree a true light version of the S7 would be amazeballs, like a 2 th unit that uses 500 watts or so and then we could use all our old PSUs that are collecting dust.

The Antminer  S8? What about the Antminer S6

The Antminer  S6 should be like the Antminer S2, S4 so a bigger unit with more hashing power and higher shipping cost.

they should think of the little guys like us who mine at home and can't afford those expensive S7's !
they should make an S7 home edition.which has fans that are really quiet and doesn't consume alot of wattage that hashes around 1.5/2 Ths and is affordable :-P

Well they could tell but you know its China so they would just cut your throat if you leak info.  We will never know these deep Chinese secrets.

It makes me wonder if their chips are varying a lot on quality.  Where each batch they make it's pretty random on what quality of chip comes out.  There has to be something as they sure are not changing speed just for fun of it.

But we likely will never know as most of the information like that will be locked tight under NDA and anyone who knows chances are can't tell.

The way things are going the S7 will be obsolete in 6 months or less the way difficulty is going, so ya something twice as efficient will have to come out.  I think a 10 th unit that consumes 1200 watts or a little less could be a prediction.  Who knows what will really happen and in what time frame but it would have to be before the halving to even keep us going in the future.

Yes, how can the latter batch be less efficient?

I don't understand it.

interesting to think the s-7 is only 60 day old and we are talking s-8

the s-7+  may have 4 s-7 cases  slide them together use 2 boards each with 2 heatsinks each.  8 boards slight downclock  = 8 x 1500 or 12th  but .2 watts or 2400 watts.

One thing to keep in mind was S5+ was pretty short amount of time and one batch.  There were S7 rumors all around during the release so we had a good guess it would be soon. 

But they can switch pretty quick as we have seen.   I mean really every batch of S7's is a question on speed at this point.   So I still think we have a while with it as their new product.

If we make it to July as OgNasty mentioned I would be thrilled.  I hope it's around then when they move to next gen I want to ride out this gen as long as we can.

I'm guessing (hoping) we don't see an S8 until July of 2016.  I would expect we'd see something like 8TH/s @ 1250 watts.

S8 will almost certainly be a 14nm or 10nm design, otherwise it doesnt have a hope in hell of competing after the halving.

We will probably see an S7+ first, but who knows what that will actually be, will it be a bunch of S7 strapped together (like S5+), or will it be a more stable and efficient s7 (like s3+).

True knowing bitcoin they will first launch there S7 +, even though the hashrate of the last batches were different

Batch 8  Hash Rate: 4.73 TH/s ±5%
Batch 7 Hash Rate: 5.06 TH/s ±5%

I think it is likely they will lead to the S7+ eventually.  I think they need to get a design down first and stick with it on speed.   As of right now every batch seems to be a new speed.... which is interesting and odd.

Once they get it finnaly down.  I think it is likely a S7+.   It might take the B-11 coming out.  They show it with different speeds up to 10T (Hashrate: 3TH/s, 5TH/s or 10TH/s options).   If they come out with that I think there then would be a push for S7+.  But I think bitmain will need a little push to do it.

First there should be S7+, actually no big reason how it is called, the best thing would be that the new miner could offer even greater efficienfy.

I don't anticipate an S8 release.

 I WOULD bet on an eventual "S7+" release, once they get the bloody things working in a stable way.


 I also anticipate the Bitmain will join the 14/16nm mining chip crowd but they'll go straight to "full custom" with BitFury and KnC have claimed to have tape-out or production on rather than doing the non-custom design like the Lketc / Innosilicon chip design(s) appear to be.

Yes I think the Even numbered S series has been replaced with the + Series. So as Amph has said next logical step is the S7+, a 3 Bay S7, controller already has the connectors, should give us, dependant on what flavour of S7 we have progressed to.  At least 15GH/s.

As to S9 well who knows, probably depends on how the Samsung 14nm chip development goes? Will be interesting to see how  LECT /BW.com do with the B-Eleven?

I found this comment from Sean Walsh, Founder of BlockC in BitcoinMagazine interesting.



Rich

isn't this the same watt hash ratio of the s7 just bigger? what's the point for more density? i was thinking that they should focus on efficiency more if they want to release something new

they can do a s7+ for a bulkier miners based on the s7 board

Anyone know if they plan an S8 ?
My predict would be 2600 W  with dual PSU with around 10-11 th/s