Member Info for Phimx

Mickydj: You post

" On what evidential or professional basis are you asserting that a team of experts could not have existed ?

Micky, firstly Google "Source: Arm Inc 2023 IPO Prospectus (IBS, July 2022)" this document by ARM, and they know a bit about ASIC design, shows approximate costs of ASIC manufacture. For a 5 nm ASIC they estimate $449 Million. Yes that's correct $449 Million. Now it defies believe that a sane man contemplating production of a 5 nm ASIC would not jump on his laptop and do a bit of Googling to establish what he's dealing with !

Or, if he is employing ASIC/FPGA experts, casually ask them at tea break " what would it take to make a 5nm ASIC chaps"

I think Gardin, in a RNS. stated something like he had reserved manufacturing time with an ASIC firm.

Therefore any sane and sensible person would, I think, be astonished that Gardin did not know the cost. Apparently from subsequent RNS he persisted with the idea, even to this day. But the cost realisation eventually filtered through to him and he abandoned that particular fantasy.

Maybe I am wrong perhaps his team did say, "hey man that's an awesome cost" but he carried on making announcements of building the fastest industry leading ASIC without any solid foundation for doing so. No public attempts were made to raise money it seems.

So, I believe there were no 20 or so 'experts' because he continued to make RNS statements apparently oblivious to the cost and complexity of the project. He was also not being advised that hashing and Method C hardware did not actually require an Oracle on an ASIC. Additionally he was oblivious to the fact that throwing away 30% of your header inputs actually provides a 43% hashing rate increase, as shown by statements in several RNS.

The alternative is, there is a team of world class experts but Gardin chose to ignore their advice and continued to make claims he had very little chance of achieving.

My qualification, I can read.

Toneman:

Always concerns me when I see apparent big buying with no significant rise!

Mickeydj: You post

" You’re entitled to your inference, but dressing it up as established fact is where the problem lies."

I'm not claiming it is a fact. I think my presentation establishes my path of reasoning which leads to my conclusion. It is based upon descriptions provided by QBT, the workings of Bitcoin mining and commonsense. You agreed with my basic hashing description which established it was not essential to place the Oracle on a hashing ASIC. The whole ASIC saga itself is worthy of an in-depth study for its sheer absurdity. It basically indicated to me and others that Gardin wasn't being guided by an expert team. Gardin fails to publish any sensible scientific data which can be held to scrutiny. The only performance figures for Method C software is, i believe "north of 10%" do you expect sensible discussion from vague amateurish statements like that from

Gardin?

Your attacks on my posts revolve around, wordplay and semantics, literal meanings. definitions and phraseology. as if we are presenting evidence in a court of law. Well sorry, I'm writing colloquially and at a reasonable speed and little time to correct my text.

As you say people can decide for themselves. It's all in Gardin's hands !

Mickdj: You post

“You’re presenting inference as established fact once again, then objecting when that distinction is pointed out.”

You seem very on heavy the inference thing. From AI

" An inference is a conclusion or educated guess you make by using evidence, clues, and reasoning to figure out something that isn't stated directly,"

Yep that's me!

"

Mickdj: You post

" You are arguing that evolving technical descriptions imply incompetence."

I wrote ignorance but who cares!

I am implying it is highly unlikely there was a team of experts because not one of them corrected Gardin's ignorance of the mining process and his mistaken ASIC and ultimately failed ASIC route. You can argue it was a design choice option or whatever but what fool would go down a multi-million pound product route that was apparently unnecessary?

Mickydj: You post

" You very much do in fact use the term “covered up” in this thread below:

So I did use that term but to my credit I was using it to describe the cover up of his flaws rather than some illegal act, as I think you were trying to suggest.

Mickdj You post

" You say “Gardin has continually changed the story”, yet you haven’t demonstrated a contradiction in the RNSs, only that your interpretation has shifted as new descriptions were released"

The story from the introduction of Method C was the Oracle had to run on the hashing ASIc. I don't think that is in dispute, go read several RNS and listen to interviews. My observation is, that was incorrect. Why was that? Because Gardin didn't understand the hashing/mining process. This at odds with his admissions of having a team of around 20 experts working for him. As I have explained creating headers to be hashed by a 1 TH/s miner should be possible but the time to evaluate each new header is unknown. Note by headers I mean modification of the header template provided by the pool by changing the Merkle root (as the extra-NONCE) but before it is sent to the ASIC

I speculate that Gardin finally realised how hashing was being performed by the ASIC and realised that an Oracle on an ASIC was not necessary. To be clear hashing of the header can be divided into 3 basic stages

1. Assemble the header from the information supplied by the pool, there is only one supplied. Performed by. the control system

2. Send to ASIC and hash with 4.3 billion different NONCEs. If no result

3. Change Merkle root to create new variation of the original header, again performed by control system and repeat.

My conclusion from the above: The story has changed it changed because of ignorance rather than development but you can spin it as development if you wish. But without doubt you have to ask the question, where was the expertise when this ASIC/FPGA scheme was being developed when any amateur miner could have probably told Gardin how mining worked. It also complements the fiasco with the CGMiner software.

Mickdj: You post

" Developmental clarification is not the same thing as cover-up"

There is no mention of a cover up in my recent posts. I write one thing and you accuse me of something totally unrelated

Mickydj: You post

" Readers can judge accordingly." Let us hope so.

I'll just paste my conclusion here

Conclusion:

In a word salad, Gardin has just described the hardware method C as the software method. He claims the software method is now possible because headers are assessed and not instances but they are the same thing. The reality is Gardin didn’t understand the hashing process, and neither did I. The hardware version never needed an ASIC to operate, it had the time period whilst the 4.3 billion NONCEs were being rolled to construct a new header. For a 1 TH/s hasher that time was 8.3 ms probably sufficient for a fast microprocessor to create a new header and assess if it was worth hashing.

Gardin has continually changed the story as he has learned and covered up the flaws in his story. So much for a team of experts !"

You a tually consider this a 'technical conclusion' ? I'm assuming you think its technical because of the text

4.3 Billion NONCES - 1 TH/s hasher - 8.3 Ms and microprocessor

Bizzare. Truly Bizarre please Google these terms to aid your understanding

Mickydj: You post

" Mrhanji, i have asked phimx previously to explain on what authority and context does he write his observations"

I've looked at the LSE rules and guess what, no qualifications are required to spout on here.

Of course you could set the ball rolling and amaze us all with your AI credentials !! Please note a vivid imagination doesn't count.

Mrhanji:

I agree, ignore my posts and accept the other views here!

I think you are destined for fame and riches.

Mrhanji: You post

" And you've just proven you have none"

An interesting conclusion Mr.

Unusual trading continues ??

Mickdj: You post

" ... relies on disputed definitions, and then treats those disputes as proof of failure.

Well what do you expect when clarity and sensible factual information hasn't been forthcoming from the Chief mumbler ?

Mickdj: You post

" Speculation is being presented as forensic conclusion"

Forensic conclusion I don't think those are my words but they are obviously your words.

Speculation I guess that means drawing a conclusion from too little hard facts. No doubt we are all guilty of that. We only have the mumbled ever changing words of a one man band to draw on and a total lack of professional and scientific evidence.

Mickydk: You post

" He changed the story” is assertion, not evidence"

This isn't a case of law, it is a BB, filled with self opiniated bar room 'experts' ! I include myself.

Yes I am stating I believe Gardin has authored a changing story line during the so called R & D phase to match his increased understanding of Bitcoin mining. From Naik's 13 year old thesis to an Oracle, which detects patterns that no-one else in the World can duplicate!

Mrhanji: You post

"Phimx can you post your background or any qualifications or papers you written on this ?

Trying to assess your expertise if even have the technical knowledge or background to even given any technical opinion on this"

Why don't you assess what has been written. I think I've explained it reasonably well. My foremost qualification is the ability to read perhaps my explanations require a bit of Polish?

Mickydj: You post

" Instances refer to nonce-variant hash attempts."

I agree

From QBT web site:

"Method C AI software Oracle, which runs on the operating system of the mining rig, performs a real time assessment of different headers starting from the job received from the pool before submitting them to the hashing boards This approach avoids processing a large number headers which have very little, or no probability to generate an hash above the pool targe"

From other RNS and interviews we know some headers do not reach the ASIC they are discarded. You are saying the template is evaluated as a job lot but still it is instances are being discarded I.e. identical to the hardware method. Common sense tells us we would expect both forms of Method C to operate similarly.

The difference is Gardin said initially that an ASIC was needed because the Oracle had to match the ASIC throughput. This was incorrect, it is feasible that a microprocessor can hash a new Merkle root and evaluate the resulting altered template during the time the current Merkle root template is being hashed 4.3 billion times.

So Gardin and his team of so called experts didn't understand the process

Mickdj" You post

" The argument assumes the Oracle must compete with ASIC throughput on a like-for-like basis. That is not what Method C claims."

Incorrect, my post specifically states for a 1 TH/s miner, there is 8.6ms between each header loaded into the ASIC. Each header differs by a change in the Merkle root. The ASIC then blasts through the 2^32-bit NONCE field I.e 4.3 Billion unique NONCE values.

The Merkle root acts as the extra NONCE field and can provide many more NONCE values than the designated NONCE field

Thanks Hexam

I'll just ignore him from now on