Topic: Overt AsicBoost Released today? (Read 1730 times)

Yeah I was thinking the same thing about Inno as I reviewed the hardware list I have yesterday. While it all seems to convenient I remember there being something in the Halong thread that there was some business relationship between Halong and Inno, just can't find/remember what it was. Regardless it shouldn't have been an open door to the tech unless there was something else going on behind closed doors.

Then I was also wondering about Bitfily they also produced miners that run AB, and I'm not sure who or how they fit into that list of licensees.

Regardless I guess it's safe to say any manufacturer who think s AB will improve their efficiency will be able to integrate it into their designs in the future, or at least test it and decide. That's assuming the patent continues to get tossed.

So its been a week and a day since the S9 AB firmware was released... think they're working on the hashrate bug in the S9 firmware before released the AB firmware for the other BM1387 based machines?

Speaking of which, if you check the Little Dragon BDPL site for its list of licensees one company very obviously missing from the list is Innosilicon who are using AB in their T1+ and T2 series of miners. The companies that are listed have NO patents to their name to 'share' with others and aside from Halong nor do they produce/sell any miners. Are they shell companies owned by Inno or otherwise associated with them? Again, the BDPL covering AB is very obviously just an attempt to get access to other companies IP.

IMHO that rather points to Inno being behind Little Dragon and the AB patent...

What can I say lost in the details/lack thereof. I guess I did assume their identities as being under one umbrella, as that makes the most sense to me when I look at their overall corporate relationships. Regardless we'll see if they stick around or do anything in the future; or if they end it as an enigmatic puppet.


This I find interesting and plays into my musings above. This "game changing" patent came out of nowhere and was being used to create a consortium of companies sharing patents. Now it is being shot down as non enforceable by a company that was easily able to adapt their machines to run in the same manner. In the end it looks like an improvement that was held back and used in an attempt at corporate espionage done out in the open.

I might finally have my chance to play around with AB gear as the T9+ is supposed to have some firmware drop, really can't make the machine any worse fingers crossed.

It's nice to know that there are manufacturers out there finding ways around needing this and getting great results. This is still where I plan on looking for new gear.

Yes the chips were designed with BM's version of AB in mind. However, after the kerfuffle over covert AB Bitmain seems to have decided to hold off turning on their overt AB until the noise over AB in general settled down AND until someone else got the pools to support it.

Why do so many folks still continue to wrongly say that Halong Mining owns the AB patent? As has been said time and again including being very clearly spelled out on the Halong site the patent is owned by Little Dragon LLC. Halong is just a licensee of it. Now as to who owns Little Dragon (and for that matter Halong) is anybody's guess as neither company wants anyone to know and both have gone to great lengths to prevent anyone from knowing. The only info on Little Dragon is that they are represented by a Chicago law firm (look on the Little Dragon LLC site) and all communications go through them.

As to violating the Patent - highly doubt it though of course I expect that to be contested. However BM made mention in their blog that the Patent has already been shot down in several jurisdictions. More to the point - you cannot Patent an end result (reuse of previous hash calcs) and there are always several ways to skin a cat. Case in point, Intel/AMD cpu's, the various makes of GPU's, etc. all do the same thing, just differently but producing the same results.

Thanks for that.

I guess even with that information, you would still need another level of transparency to prove whether or not Bitmain had been using this firmware upgrade for any definitive amount of time. Just seeing blocks within their pools, or even the unknown pools can't be verified as the claim could be that it was a known asic boost miner.

With all this, I've been going back over the beginning of this thread and I admit to being lost.


How did this play into things, if the firmware upgrade works on all s9's does that mean the chips were already designed with this feature in mind?

Does this somehow violate the patent held by Halong?

Again, covert AsicBoost no longer works, because of segwit.

There is therefore no longer any miner that is able to use AsicBoost in secret. (This has been the case ever since block 481824, when segwit activated.) The non-standard version numbers of their blocks would immediately give them away.

If it benefits to mine with AB whether covert or overt, it's just bad business to not to on every piece of hardware. If this has been practice it's been a poor one.

Yes, there is.

Non-AsicBoost blocks have a version number of 0x20000000. AsicBoost blocks may have different version numbers.

For example, block 546932 — mined by ViaBTC, which currently does not support AsicBoost — has a version number of 0x20000000. Block 544704, which is a known AsicBoost block, has a version number of 0x3fffe000.

When it comes to Bitmain, who knows what goes on behind the curtain.

I want around for the "yay" Bitmain days just, the bad reputation and ill will part of their business so far.

Is there a way for the public to tell if a block was mined by an asic boost piece of equipment? I know the pool can tell but I sort of assume that's just because the hardware is connected, and you know what hardware runs asic boost.

Most big businesses are thieves and practice illegal moves all the time.  Smart ones do it in ways they do not get caught.

So they  could simply mined to an unknown address
See below

You would have seen more AB blocks solved by Antpoo and BTC and yadda yadda.

my guess would be   a 10 to 15% improvement in efficiency .

So if your s9i  was doing 94 watts a th it would do  80-85 watts a th

but  since the m10  is the same size chip and does 70 watts a th  I guess that would be possible.

Apparently its a simple firmware upgrade available now for the S9 from the Bitmain support page. Some testing going on now with an S9 on Kano to see if there is any noticeable improvement in hash rate or a reduction in power consumption.

Given the new firmware was given a "LPM" designation, I'm leaning towards it reducing power consumption for the same hash rate.

That is exactly my question.  What is the effect on power consumption and hash rate for this.  Is it worth even bothering to do given the S9's mine at a loss.  Is it really going to change anything?  Would love to hear more and what you have to do to use this feature.

Hah, I knew it.

Did they figure it out like yesterday or have they been doing that with their own machines for a long time? 

I haven't really been following this to closely, as I don't own any asic boost hardware.

Bitmain tweeted that they have new firmware to activate this for the S9's

https://twitter.com/BITMAINtech/status/1054329450018435074

What sort of hash increase can people expect? Are th eefficiency numbers going to get better for the S9's that are currently available or is this just going to be more powerful?

I’m pretty sure, that Halong didn’t design and implement a 10nm ASIC without help. Design teams capable of 10nm implementations are not coming out of nowhere.

In principle Innosilicon is a design service company. They are realizing any kind of microchips for any customer who is paying them. Normally that kind of service companies are “pure-play design service”, which means that they are not releasing products labelled with their company name. Innosilicon seems not to be so strict here, they have own products too.

However, I guess that Halong contracted/paid Innosilicon to implement their 10nm ASIC as a turnkey service. Halong covered the complete NRE costs, took all the risks and is the owner of the resulting ASIC design and mask-set. The engineering was done by Inno (maybe Halong developed the ASIC architecture and parts of the front-end design).
That is the way it goes, nothing really wrong here.

Could Innosilicon have created Halong to avoid sharing its patents with BDPL?

There are strong indications that they are related in some ways. The following post summarizes the key points:
https://bitcointalksearch.org/topic/m.33551682

How can a startup like Halong afford to create prototype device? A mask is a multi-million dollar investment. Is it possible they acquired their chips from another manufacturer or that Halong is actually Innosilicon? My bet is on the latter.

If Innosilicon is supplying chips to Halong, shouldn’t Innosilicon also be obliged to join BDPL if they manufacture chips with the AsicBoost design?