Will Bitmain Antminers S9 and T9 be obsolete if Segwit is launced ?

[Xircom]

Rumors says that code of ASICBOOST has been discovered at Antpool.
If it is the case that Bitmain has secretly incorporated ASICBOOST into the T9 and S9`s they will be more or less worthless if segwit is launched and thats why Bitmain is fighting like maniacs to use Bitcoin Unlimited.
Can anyone reveal what hasrate a "real" S9 and T9 then will be putting out on the segwit network without cheat?
Do we have to take 20-30% of present hashrate pr unit?

[1714844064]

1714844064

[1714844064]

1714844064

[1714844064]

1714844064

[1714844064]

1714844064

[1714844064]

1714844064

[1714844064]

1714844064

[HagssFIN]

Let's take a 12.93Th/s S9 for example.
You mine it at Slush or Kano pool for example.
There is no support for the Asicboost.
You get the 12.93Th/s hash rate.

There would be no 20-30% drop to our current hash rate.

Also there is not yet proper proof of the Asicboost being used in the main Bitcoin network.
The Asicboost method has been proved to exist, for sure, and that Bitmain is involved in that with bmminer configurations.
But we can't be exactly sure if it has being used yet in the main Bitcoin network.

[Xircom]

So are you saying if those same miners are pointed at Antpool, you will have a gain over other vendors ?

[HagssFIN]

Until there is no proof of gain being reached with Asicboost, I would say there is no gain in the current Bitcoin network.
I would guess, and now this only a guess, that being able to use Asicboost would require a change in the Bitcoin protocol.
This could have being proposed in the Bitcoin Unlimited protocol and would explain Bitmain's strong support on the BU.

[Rabinovitch]

So are you saying if those same miners are pointed at Antpool, you will have a gain over other vendors ?

I guess if those same miners are pointed at Antpool by BITMAIN, they will have a gain over other Bitmain's asic miners owners...

[adaseb]

That Sam Cole guy (KNC Miner) said that the only efficient way to use ASICBOOST is with a custom ASIC chip (Different from S7/S9). Since it would use less wafer space. And it would be completely useless if Segwit is launched.

However there is no proof if these actually exist.

[HyperMega]

That Sam Cole guy (KNC Miner) said that the only efficient way to use ASICBOOST is with a custom ASIC chip (Different from S7/S9). Since it would use less wafer space. And it would be completely useless if Segwit is launched.

However there is no proof if these actually exist.

Timo Hanke’s asicboost paper explains the benefits and hardware prerequisites quite well. If I understood it right, there is no need for changing the Bitcoin protocol to enable it, it just has to be designed in the ASIC and one would need a custom miner software. So in theory all S9 could already run with asicboost and nobody would know.
But the paper also shows, that if you want the full asicboost potential in terms of silicon area and energy efficiency, the resulting ASIC could only be used with asicboost.

So if Segwit would prohibit asicboost (I don’t know if this is really the case), these kind of ASICs would be garbage.

I also guess that if somebody implemented asicboost in their ASIC (most likely anybody with <0.10 J/GH ) they probably added a kind of a “legacy” mode, which would enable a classic operation mode again, in case that somebody changes the protocol. This would cost silicon area, but still saves a lot of power.
But maybe they were just too greedy to add this backup solution.

[Xircom]

I just noticed a frase in Bitmain`s statement from last night re. ASICBOOST.

One line really got my attention:





So, that line tells they have tested or seen ASICBOOST on main net, but not quite as the patented way......
They also states they want to have the advantage of ASICBOOST in China for their hosting facilities....Am I the only one feeling this knowledge should have been out to the masses so the network would grow and transactions faster ?

[fanatic26]

I feel like people have no concept of what segwit or asicboost actually is....they just see some reddit post demonizing it and fall in line when they really dont have a clue what it does or how it affects things. Everyone is so focused on calling bitmain the devil they dont care about the truth. Its like burning someone for being a witch back in the day...there was no basis in reality for any of the claims levied against people yet they still burned because the unintelligent masses would gather in mindless mobs and parrot what one or two loudmouths would say until everyone believed it to be true.

[NotFuzzyWarm]

I feel like people have no concept of what segwit or asicboost actually is....they just see some reddit post demonizing it and fall in line when they really dont have a clue what it does or how it affects things. Everyone is so focused on calling bitmain the devil they dont care about the truth. Its like burning someone for being a witch back in the day...there was no basis in reality for any of the claims levied against people yet they still burned because the unintelligent masses would gather in mindless mobs and parrot what one or two loudmouths would say until everyone believed it to be true.

Ditto and also +1
It is the proverbial Tempest in a Teapot and much adieu' about nothing.
Please let this subject just die...

[kano]

I feel like people have no concept of what segwit or asicboost actually is....they just see some reddit post demonizing it and fall in line when they really dont have a clue what it does or how it affects things. Everyone is so focused on calling bitmain the devil they dont care about the truth. Its like burning someone for being a witch back in the day...there was no basis in reality for any of the claims levied against people yet they still burned because the unintelligent masses would gather in mindless mobs and parrot what one or two loudmouths would say until everyone believed it to be true.

+1

Lulz, so since joining Bitmain, you've been dumbed down to a '+1' level poster who knows nothing about what his company is actually doing?
Why not provide some information? Or you don't know what your company is doing so you shouldn't be posting at all?
You should also update your sig, you don't work for spondoolies

The original firmware of the early S9, before they were sold to the public, included a change to the stratum protocol to support the document I posted in the S9 thread.
https://bitcointalk.org/index.php?topic=1493601.msg15634328#msg15634328

The S9 didn't work on ckpool due to a stratum connection mode incompatibility.
Yes pre-sale S9s were tested on ckpool.
They 'modified' the stratum change so it works without failing to connect to ckpool.

Of course as mentioned above, this is verified here also:
https://blog.bitmain.com/en/regarding-recent-allegations-smear-campaigns/

If they were using the 'Overt' method, you'd see it in the block headers.
If they were using the 'Covert' method, it would mean a power saving, so not a direct hash rate increase of each miner, but an overall hash rate increase due to more available power.

Most likely they are using the 'Covert' method.
Their blog pretty much implies it by not denying it, and confirming it's in their miners, and saying they have paid for the rights for it in china.

[NotFuzzyWarm]

I'll agree with kano aside from one point, re:

and saying they have paid for the rights for it in china.

Actually Bitmain said "they have the patent on it in China". Emphasis to 'in China' is required. If the Published Authors behind ASICBOOST did not file for copyright/patent protection in China or for that matter most any other specific country then they are Boned. Legally Bitmain is well within their right to claim it as theirs (in China) and do note that to cover their ass they specifically mention their 'right to use in in their mining operations located in China.

In other words, the logic may be present in the ASICs but to minimize Export issues (patent/copyright infringement) it does not run on miners outside of China (so far they say). Not saying that it is right -- just that that is how International law views it.

Been there, done that. Filing for International patent coverage is a Bitch and a half... At least THAT is one thing that the EU made easier for those concerned. File once and covered for all member States. Asia OTH....

[Xircom]

NoFuzzy, that was a quick turn around 
I can confirm that Kano did discover this with the first batch of S9, I was at that time pointing my S9's at kanopool and we did have a lot of discussion about the behaviour of the S9.
Anyway, I feel it's a cheat and I think the only way Bitmain can get a slight trust back is to open this feature for everyone who owns Bitmain miners.

[NotFuzzyWarm]

Xir, you miss my point: this kerfuffle over asicboost is IMHO pretty pointless. According to the Bitmain blog 'others' (Avalon/BitFury/BW?) also have the logic in their chips. That I have no reason to doubt. Thing is, if AB was/is so great, it would be in use by all the miner makers and after a blip in diff everyone would be mining at the same level again.

[Guy Corem]

https://medium.com/@vcorem/the-real-savings-from-asicboost-to-bitmaintech-ff265c2d305b

[NotFuzzyWarm]

https://medium.com/@vcorem/the-real-savings-from-asicboost-to-bitmaintech-ff265c2d305b

Thanks for the info -- bookmarked for future reference
I wonder if Bitmain's Chinese Patented method of using Merkle Collisions (I do like that name better...) lists your work and that of the ASICBoost Authors as Prior Art? I would hope that the Chinese Patent process would at least require mention of that...

Rather helps confirm my feelings about AB not being worth the fuss given it. thx

#%(*^&#! There I go falling for this freakin' thread again... 

[Guy Corem]

I wonder if Bitmain's Chinese Patented method of using Merkle Collisions (I do like that name better...) lists your work and that of the ASICBoost Authors as Prior Art? I would hope that the Chinese Patent process would at least require mention of that...

I don't have any information about Bitmain's patent, except rumors that it is copy of ASICBoost.

[NotFuzzyWarm]

I wonder if Bitmain's Chinese Patented method of using Merkle Collisions (I do like that name better...) lists your work and that of the ASICBoost Authors as Prior Art? I would hope that the Chinese Patent process would at least require mention of that...

I don't have any information about Bitmain's patent, except rumors that it is copy of ASICBoost.

Found it... Surprised no one else bothered to Google it...
https://www.google.com/patents/WO2015077378A1?cl=en
The fact that it lists Timo Tobias HANKE, Sergio Demian Lerner as the Authors suggests to me that Bitmain bought the rights to it (for in China). Hmm. The Plot thickens.
Filed Nov 19, 2014

[Guy Corem]

I wonder if Bitmain's Chinese Patented method of using Merkle Collisions (I do like that name better...) lists your work and that of the ASICBoost Authors as Prior Art? I would hope that the Chinese Patent process would at least require mention of that...

I don't have any information about Bitmain's patent, except rumors that it is copy of ASICBoost.

Found it... Surprised no one else bothered to Google it...
https://www.google.com/patents/WO2015077378A1?cl=en
The fact that it lists Timo Tobias HANKE, Sergio Demian Lerner as the Authors suggests to me that Bitmain bought the rights to it (for in China). Hmm. The Plot thickens.
Filed Nov 19, 2014

What you found is Timo & Sergio original patent. Not Bitmain's patent.
The important date is the priority date: Nov 19, 2013
The date the provisional patent was filed.

[NotFuzzyWarm]

Ok, this one references Bitmain directly filed in the EU
https://worldwide.espacenet.com/publicationDetails/biblio?II=30&ND=3&adjacent=true&locale=en_EP&FT=D&date=20160113&CC=CN&NR=105245327A&KC=A#
Original document https://worldwide.espacenet.com/publicationDetails/originalDocument?FT=D&date=20170302&DB=&locale=en_EP&CC=WO&NR=2017031967A1&KC=A1&ND=4#
Interesting that the date of original doc is Feb. 2017...

However, Bitmains patent translated from Chinese https://www.google.com/patents/CN105245327A?cl=en has a Priority date of Aug. 2015.