Member Info for Phimx

Usersteve:

You could slow the clock rate of your Oracle ASIC but you lose the advantage of finding a winning hash. I cannot see that being a practical way to operate

Usersteve:

So I'll admit one system found a bit coin but how much of an electric bill did each pay ? I say exactly the same! What if the other system just happened to find the winner, again both electric bills the same but the Oracle system processed more inputs and will undoubtedly have a better probability of finding a winning hash

In fact the more I think about it the more I believe there is no 30% energy savings except as I've already said. It only happens for unusualand limited hashing regimes.

Think about 2 identical systems, one with an Oracle that consumes virtually no power. When will you save 30% on your electric bill ?

Love to hear Dooogle, mickeydj or Koolkat answer this question but I doubt very much they will!

Awl007:

T here is a possible saving of 30% to be had in electric costs but this only realised if two equal systems hash an exactly the same amount of inputs, one system with the Oracle. Reality is the hash ASICs works non stop but the Oracle system would get through 43% more inputs for the same energy.

I think 30% energy saving only is available under specific conditions. The 43% gain is what most hashers would want, this comes with no additional energy cost

In fact is there any lab data at all ? If a team of professors and experts are conducting testing the it is impossible to believe that not a single person failed to notice the results were better than expected by 13%

Not a single member of the team !

Thus the conclusion is the test data or scientific evidence to back up the claims just isn't there!!

Awl007:

Read this carefully then you will understand - hopefully

" The material competitive advantage in mining enabled by the AI Oracle may be achieved either by (i) reducing the energy cost of mining by approximately 30%; or (ii), accelerating the mining speed at current energy consumption and costs with approximately a 30% greater hash rate."

This is incorrectly, using the same energy as a non-oracle hasher will give you, all other things being equal, a 43% greater hashing rate.

Not as single person here has attempted to refute this, how could they it is trivial to obtain. So if Gardin has loads of irrefutable data as once claimed how does that data show a 30% gain.

The logical answer Gardin didn't think to hard when he was writing his story and he doesn't have lab data.

Usersteve:

"The Company has decided to facilitate this development with a mature process node chip due to its lower manufacturing costs rather than for example, a market-leading 5nm prototype ASIC, which would cost nearly £2m. "

If this were true I.e. QBT are going to use a mature process I.e. a slower chip they would simply defeat the purpose of having a faster chip. The slower chip would negate the faster hashing algoritm.

Just drivel to fool the gullible and keep the money coming in.

Usersteve:

Well done for wading through the drivel from Gardin. I'm sure he makes up the plan as he goes along surely there can be no experts advising him.

Usersteve:

For heavens sake that was 3 years ago, Gardin has long forgotten that little teaser story he wrote.!

AND 2 nm are the in thing now

Come on Gardin tell everyone how you calculate efficiency at 30% when it is 43% ????

Usersteve:

As you've shown you can't trust what Gardin writes. He has actually shelved the 5nm ASIC to concentrate on the other avenues, it's in a RNS somewhere. The whole RNS story is not easy to follow, there obviously isn't any planning involved, I'm sure he makes it up as he goes along after reading comments here.

How is he going to explain the 30% efficiency reports from his huge lab data and his team of experts when a trivially simple sum tells it should be 43%, based upon discarding 30% of no-hope inputs ????

Deathly silence from everyone on that very important topic, only Koolkat attempt the calculation with his "program" Surely he was joking ?

Dracula1: You post

" Phimx, are you neglecting your other shares. Do you not think they may need your expertise."

I have no expertise, some commonsense perhaps and an ability to Google.

Usersteve:

Of course there will be no ASIC chip!

Although Gardin had plenty RNS pushing what QBT were going to do in the ASIC field it was all just bluster and drivel. He didn't realise how expensive it would be. Of course if he actually had a team of experts presumably a major obstacle like that they may have remembered !

Koolkat: You post

" You can't blame the Chip company for taking their time to determine that Method C is the right software to improve their chips."

Method C 𝗜𝘀 𝗻𝗼𝘁 𝘀𝗼𝗳𝘁𝘄𝗮𝗿𝗲

Drutor: You posted

" what happend to this media video they were making."

What happened to the SaS application ???

Drutor:

As I have previously stated, yes they have the 2013 M.Sc. optimisation work by Naik, he also acted as a consultant after for QBT.

Yes they have a project, personally I think it is outdated and the optimisations are old hat 𝘁𝗵𝗲𝗿𝗲𝗳𝗼𝗿𝗲 it is very unlikely to have any commercial worth. This is why Gardin continues to attend these conferences and produce RNS which say nothing of substance. As my previous posts show he has claimed to have been in discussion with 2 big Chinese and 2 big US players over 2 years ago. Discussions, development and drivel.

How could all the testing and the huge amount of test data he is supposed to have amassed not show a 43% hashing increase for his Oracle ???

Dooogle:

Here's more from Gardin!

" The Company’s board of directors believe this is a significant milestone, since this proprietary technology has been used in trials to mine Bitcoin with a competitive advantage against the same hardware without the AI Oracle implementation. The material competitive advantage in mining enabled by the AI Oracle may be achieved either by (i) reducing the energy cost of mining by approximately 30%; or (ii), accelerating the mining speed at current energy consumption and costs with approximately a 30% greater hash rate."

Care to comment as to why Garden and his team of world class experts don't understand why rejecting 30% of inputs allows for a 43% improvement in hashing rate, particularly after irrefutable results from laboratory testing.

Come on Dooogle copy and paste something

In the Courtois paper https://arxiv.org/abs/1310.7935 submitted 29 August 2013 i.e.

The Unreasonable Fundamental Incertitudes Behind Bitcoin Mining

Is the statement:

“In this paper we are the first to develop such techniques openly and publish them. We have invented these techniques independently from scratch and to the best of our knowledge they are free of any intellectual property rights. However we expect that ASIC designers have already done similar optimizations and some of these techniques could have been patented.”

So it is reasonable to expect that time has not stood still in the last 12 years and ASIC designers have incorporated many optimisations into their ASICs. Further, it is also reasonable to expect that the optimisations of Courtois have been fully explored and exploited by the Bitcoin industry over that time. Note QBT’s first patent application, ASIC UltraBoost is based upon those optimisations. The UltraBoost patent has 𝗯𝗲𝗲𝗻 𝗿𝗲𝗷𝗲𝗰𝘁𝗲𝗱 because of prior art, i.e. the work had been published well before the patent application.

The industry isn’t interested because QBT’s project is old hat

Dooogle:

You are fond of post drivel from Gardin, let's remind ourselves of other great proclamations he's made.

" The Company is working with two independent partners to conduct verification tests on the final performance of QBT’s prototype ASIC. Their role is also to advise on how best to achieve efficiencies in coding, in order to best exploit the specific manufacturing technologies of individual silicon fabricators, when it comes to manufacturing of the prototype chips."

3 long years ago, Dooogle remind us all how that turned into fresh air !!!

Like much of what Gardin says !

Would be nice to see so.e American money flowing into IES to support the US progress ?