Member Info for Phimx

So, according to AI

For common, albeit high, pool difficulty) I.e. 2^60 you would expect:

2^36 shares I.e. quality hashes

So the Oracle is going to provide you with 2^36 possible hashes

Then what ????

This bit is a real cracker

" Further, the software version of the AI Oracle is able to assess the entire `job' received by the mining pool and its allowed variants (also called the meta-header), rather than individually assessing single bits of various key regions of the header"

So what I think Gardin is saying here, because the software cannot possibly match the speed of the ASIC, the entire header and all possible NONCE variations are assessed in one go. The Oracle then tells the ASIC which NONCEs to hash.

Truly unbelievable, without performing any assessment of the billions of NONCEs the Oracle can predict quality hashes. Since all legitimate NONCEs are the same for every block then the Oracle looks just at the block information and makes a remarkable and in the reat of the world's eyes unbelievable prediction.

Consider this with an 8 byte extra NONCE i.e. a legitimate NONCE variation, there are 2^96 possible NONCEs, at 1 TH/s hashing rate it would take 2510 years to compute all hashes.

So a tiny logic circuit i.e. 1 to 4 % of a single double hash circuit, selects all the quality hashes from 2^96 possibilities. How many 'quality hashes will there be" ?

𝗣𝘂𝗹𝗹 𝘁𝗵𝗲 𝗼𝘁𝗵𝗲𝗿 𝗼𝗻𝗲

Anyone care to decode Gardin's 3 step process

" Data collection of the ASIC Hashing behavior, to train the AI Oracle Integration of the AI Oracle with the existing mining rig's software environment Testing and fine tuning of AI Oracle's performance during live mining."

ASIC hashing behaviour ??

the AI Oracle Integration of the AI Oracle - a wonderous piece of nonsense!

Not a single fullstop in the description of the 3 step process. It looks like a cut and paste from an email. Certainly not a quality description from world class researchers.

Oldsharedog:

Your maturity and wisdom shines through the gloom. An inspiration for everyone to aspire to !!

0134:

Excellent it's easy to see where Micky gets his lines from !!

Did Gardin go to NEMS 2026 ?

RNS 15 November 2025

"... the performance of this new software-only version is 𝗶𝗻𝗱𝗲𝗽𝗲𝗻𝗱𝗲𝗻𝘁 of the ASIC being used in the miner."

𝗣𝗲𝗿𝗵𝗮𝗽𝘀 𝗻𝗼𝘁 𝘀𝗼 𝗶𝗻𝗱𝗲𝗽𝗲𝗻𝗱𝗲𝗻𝘁 𝗮𝗳𝘁𝗲𝗿 𝗮𝗹𝗹 𝗯𝗲𝗰𝗮𝘂𝘀𝗲 𝗻𝗼𝘄 𝗚𝗮𝗿𝗱𝗶𝗻 𝘀𝘁𝗮𝘁𝗲𝘀

RNS 6 March 2026

" QBT's R&D team has begun to analyse the rig's unique software code and the ASIC's architecture in order to implement the porting of the software version of Method C AI Oracle onto the mining rig's operating system as soon as possible."

A carefully crafted RNS, written to mitigate all criticisms raised on the QBTs chat boards. So many words written providing so little detail.

I'm going to sit back and watch everyone lose their money

Here comes the placing !

Beekeeper pushing a 13 year old optimisation technique, that's been well known by the industry for that length time and suggesting its worth a billion pound.

Haha

Oldsharedog: You post

" They are NOT caveats you stupid boy - they are words generally found in the Oxford English Dictionary,"

The beauty of words is you string them together to create something meaningful, far beyond the dictionary definition.

You display pompous ignorance Oldsharedog

Oldsharedog: You post

" Does anyone give a toss about these so called arguments 🤷‍♂️"

So if no one gives a toss then thers nothing to discuss, is there!

Typically Gardin does not specify what the 8-12% I.provement is with respect to ?

If it is the standard SHA256 implementation then I would expect all the big miners, with their own ASICs have already well optimised their ASICs to better the standard. Naik's optimisations have been public for 13 years so they are no secret and if of benefit would have been employed in ASICs many years ago.

Note also, an ASICs performance above the standard will be proprietary info and highly unlikely to be public information therefore Garding is very probably using the standard as a reference.

Mickdj: You post

" But jumping from “no confirmation yet” to “testing must be imaginary” is not a logical conclusion, and that process in itself is imaginary .

You have deliberately mislead readers here and tried to discredit what I have wrote. I wrote

𝗦𝗼𝗺𝗲𝗼𝗻𝗲 𝗵𝗮𝘀 𝗶𝗺𝗮𝗴𝗶𝗻𝗲𝗱 𝘁𝗲𝘀𝘁𝗶𝗻𝗴 𝗶𝘀 𝘁𝗮𝗸𝗶𝗻𝗴 𝗽𝗹𝗮𝗰𝗲

With reference to the post by Chipofftheblock

Nowhere did I write " testing must be imaginary"

Mickydj: You post

" What we don’t yet have is an RNS confirming completion of testing or commercial adoption"

So after after confirming exactly what was know I.e. the NDAs need to provide the gear! You come out with the line above. BUT it should be

What we don't yet have is RNS confirmation that the equipment was delivered or software was installed without problems or testing by the NDAs has actually begun.

AND that is all a long way from commercial adoption.

Let's keep it sensible Micky, lea e out the wishful thinking!

Oldsharedog: You post

" I'm very disappointed to find a lack of imagination from the cutting crew this morning - maybe they feel as though the release of big news is imminent and are sulking."

But the RNS today said very little! Possibly it may suggest Gardin could be in a nursing home by the time any result comes from the patent office!

Oldsharedog: You post

" I don't blame them you only have to go back a few years to see their unwavering determination to rubbish QBT - only to be potentially caught out years down the line with an unarguable positive that is about to land - what a waste of time they have delivered on themselves - but it has given some of us endless hours of unbridled laughter along the way :)"

That is a vivid imagination you have there!

But who am I to dispute the PlanV guys in Munich

Google

" single sha256 channel hardware requirements"

From that same RNS

" We have determined that the AI Oracle will not affect the performance of any given SHA-256 architecture of an ASIC chip. All three requirements have been met by our FPGA/ASIC design team. This has been an outstanding result for the team and QBT. Moreover, the AI Oracle implementation overhead between approximately 1% and 4% of a double SHA-256 lane, is an extremely impressive result."

The amount of logic required for the Oracle is indeed minute 1 - 4% of a double SHA lane! The Oracle must consist of a few logic gates and a register or two !!

Difficult to believe a circuit so simple has not been implemented by others !

Usersteve: RNS 15 January 2025

“While the patent application is meant to protect these innovative applications of the AI Oracle implementation, the core technology for the generation of the Oracle, i.e., parameters to build the AI Oracle generated by the Machine Learning Model C have been kept separate as these represent an asset for the Company which it intends to protect as an industrial secret.”

Mickdj: You post

" The source is the company’s disclosures about Method C testing with ASIC partners."

But exactly where is this stated. But first let's avoid confusion, as far as I know Method C hardware cannot be tested because it has not been incorporated in an ASIC, in fact that project has been suspended until phase 2, what ever that may be.

So we must be dis using Method C software and we all know the recent statement on that was RNS 5 December 2025 that said:

"The process has reached the stage where these three companies will now be needing to make their mining rigs available to QBT's testing team at its laboratory in Milan, with access to their relevant source code. This will allow QBT to install the software version of Method C AI Oracle ..."

Followed by the statement 4 February 2026

" The gathering saw key representatives of the Bitcoin industry attending as participants, allowing QBT’s management team to engage with those ASIC manufacturers which have signed NDAs with the Company. These discussions were focused on the timing for the delivery of both hardware and software from these potential customers and commercial partners"

So from the information by QBT, there is nothing to suggest hardware has been delivered to allow software to be installed. So the sensible conclusion would be

Someone has imagined testing is taking place ?!?