Hello again and thank you for the reply.
A couple of things to notice.
Whilst I understand that the US patent office has a bit of a reputation for allowing any old rubbish through that might not apply in this particular instance. Also this application is not for QBT's fantasy AI pattern spotting magic beans. It is a legitimate range of hardware optimizations based on real engineering.
Perhaps you might invite -ck and Kano to take a look.
In terms of the patent process. As you have spotted from the dossiers I have linked to the subject matter is being prosecuted in various jurisdictions. In this case under what is known as the PCT, Patent Cooperation Treaty route. You apply first in your country of Origin, for QBT this is the UK. This gives you a priority date.
https://en.wikipedia.org/wiki/Patent_Cooperation_TreatyWithin 12 months of that filing date you are allowed to transfer your application to the PCT route which allows you gain protection for your idea across the World in countries that are part of the PCT. This includes the US. Yes QBT are having problems in the EU under EPO, European Patent Organization, but they are the same as those they had in the US which, after a bit of back and forth, was eventually granted.
It is likely that this will ultimately happen under EPO and elsewhere. They will gain similar but limited of protection and be able to make a pain of themselves in other countries.
Again these are valid hardware optimizations similar in some ways to what was done with Covert and Overt ASICBoost as invented by Timo Hanke.
https://arxiv.org/abs/1604.00575This time the optimizations have been developed by Rahul Naik, now working for QBT, previously a student of Nicolas Courtois at the LSE, London School of Economics, as an extension of that work,
https://www.academia.edu/82035678/Optimizing_SHA256_in_Bitcoin_MiningUnfortunately a lot of this stuff is behind paywalls and logins and I cannot attach files here. On the other hand this stuff might bend your head. However as I suggest the work is legitimate rather than being a pie in the sky punt on SHA256 having patterns.
If you refer to the original specification,
https://register.epo.org/ipApplication?documentId=M013HDNGWFYBX99&number=US.202218696073.A&patentScope=falsethen you will find a number of sections starting on Page 25) of 46) of the PDF dealing with each SHA256 block in the Bitcoin Mining Implementation. SHA256(0), SHA256(1) and SHA256(2), these include tables for the Message Scheduler Calculations that are classified in the right hand column as being variously Incremented, Stored, Partially Stored, Always Zero or Not Optimized. This is where the savings claimed lie.
There are summary tables at the end of each where the gate savings and/or operation savings are listed.
As initially presented QBT tried to claim that all of these were novel and therefore subject to protection. The US examiner determined that this was not the case based on prior art including the thesis written by Rahul Naik and placed a restriction on what was allowed in the independent claims.
The basis of that restriction can be found in the text beginning on Page 21) of 46) which describes Figure 3) in the drawings. It applies to W[t] for t = 16 to 63 of the 64 rounds in SHA256.
It is my conjecture that many of the optimizations proposed, at least in the dependent claims, will not be fully protected and might be implemented avoiding the restrictions in the independent claims. Furthermore it is possible that others have already implemented these optimizations but have kept them as 'secrets'.
In such cases there is the distinct possibility that QBT is going to Patent Troll them as hinted at by their post on X,
https://x.com/QBTOfficial256/status/2093359908485730552"That's a nice ASIC you've got there. We think you may be violating our granted Intellectual Property. Perhaps you would like to pay us some money?"
Either those challenged will fold and pay the ransom or they will be forced to litigate. I am suggesting that they should act collectively and be preemptive. After all QBT are firing the warning shots but according to me they are either blanks or are a fair few grains short of a full charge.
It is at least in part a bluff and I hope that others with more knowledge than myself will challenge that bluff and make that challenge publicly known.
To date QBT have survived despite achieving nothing by regular placings on the AIM market that by and large have been supported by naive Private Investors. They are due another placing and I have no doubt that they will be using this granted patent and a pretense that it has more value than deeper inspection would suggest to access such investment.
