Member Info for DSFLAT

“They could get lucky once, twice or even three times within a short period of time”

No, they couldn’t - the odds would be in the trillions of doing it three times . Even if they did it once it would just be put down to absolutely unbelievable luck - especially as they are only touting a 10% improvement.

FG has been known to tell a few porky pies over the years, solo mine 1 BCH block, provide proof of that achievement and divert to the other blocks were successfully mined via a large BCH mining pool.

Miraculously mined by the dusty S9 but unfortunately due to the technicalities involved it is not possible to explain the evidence of their existence ;-)

If Davis Lenigas were FG its highly likely that's what he'd be doing

They could get lucky once, twice or even three times within a short period of time...

Why they delayed actually mining this long is a mystery to me, they even started out with a mining Burger Van conversion and proclaimed a new type of eco mining facility

Now they are down to one dusty old S9 and testing a black box (Proto rig)

Agreed, but they could get lucky...

Showing proof to a rig manufacturer or an after market optimization company would require a prolonged set of test results

Providing the market with a reason to go All In Loco LaLa Land could be as simply as getting lucky within minutes, hours or days

My guess is QBT has stripped the old S9 down to see what's in it and much like Don Pettit the NASA astronaut said -'we used to have the technology to go to the moon but we destroyed that technology and it's a painful process to build it back again'

QBT could buy more S9's at around £100 a piece as FG said, but after goofing around with millions of pounds for years do they have that kind of cash to spend on another S9 is the question ;-)

Antminer S9 has a hashrate of approximately 13.5 TH/s. As of current network conditions, the Bitcoin Cash (BCH) network hashrate is around 6.36 EH/s (6,360,000 TH/s), and it produces a new block roughly every 10 minutes.

Using the formula:

Probability per block = (Your Hashrate / Network Hashrate)

→ 13.5 TH/s ÷ 6,360,000 TH/s ≈ 2.12 × 10⁻⁶ (or 0.00000212% chance per block).

Since one block is mined every 10 minutes, the daily probability of finding a block is roughly 0.0005%, or about 1 in 190,000 chances per day.

Dust off the old S9, beat those odds and perhaps QBT can report something of substance :-)

Maybe things are starting to come together and FG is feeling very positive about the outcome

I was going to bring up this subject today, it's an important topic and worth discussing imo

Hexam wrote 'Even before all that they’ve actually got to PROVE it works without actually mining BTC - something which nobody has explained how they can (including QBT).'

This where I believe solo mining or even pool mining Bitcoin Cash comes in.

Finding winning blocks on BCH is considerably easier than on BTC due to the lower hashrate

Whilst not quite the same the winning hashes are on SHA-256 and given the abundance of AI compute available now it shouldn't be too difficult to extrapolate BCH mining results into potential BTC mining results

BCH is slightly more profitable to mine but the majority of mining hash remains on BTC at this time, but that can and should change in the future (Big blockers expect a migration from small blocks to bigger blocks as the mining reward halves into dust and the transaction fees fail to support the miners)

That's how I think QBT will have to prove up their current optimizations which could lead to new findings and better optimizations

An old S9 would suffice in initial BCH mining, a fleet would be better and achievable if the SP gets an insane pump where a placing could start QBT on their own optimized mining journey (like they said a few years ago ;)

IF QBT far exceed even their own mining optimisation expectations and report that news to the market the SP could easily run away into LaLa Land, as often happens with this sector

I maintain that solo mining Bitcoin Cash could deliver results quickly and be the catalyst to send QBT price on a big run

Commencement of Testing of the AI Oracle on a Bitmain Mining Rig

Quantum Blockchain Technologies plc (AIM: QBT), the AIM-listed investment company focused principally on a research, development and investment programme within blockchain technology is pleased to confirm that an internal project, which it started in August 2025, to use a Bitmain Antminer S9 ("S9") mining rig as a testing and commercial demonstration environment for the software version of QBT's Method C's AI Oracle, has now reached the live in-house testing stage.

Expect some QBT solo mined Bitcoin Cash blocks in the near future is my prediction, QBT solo mined BCH blocks far exceeding the expected rate for a single S9 and QBT should be off to the races...

I've been digging into BCH mining today which showed the same winning hash characteristics as I put forward on here last week

Coincidentally this morning I was looking at S9 mining rig availability in Thailand

'In order to expedite the development, the ASIC manufacturer has provided QBT with an MDK (Mining Development Kit)'

Block claims Proto is the first time an MDK has been combined in a mining package, discussed 10 mins into the video

Tatum the large scale mining project lead also questions the number of Proto power supplies

Compass Mining Aug 2025

In this episode of the Compass Mining Podcast, we’re joined by Tatum Cave, Compass Mining’s large-scale project lead and safety overseer at our Denver Service Center. With years of experience repairing, deploying, and stress-testing ASICs, Tatum shares his honest perspective on what Proto Rig could mean for miners in the field.

Proto Rig: How Block’s New Mining System Could Transform Bitcoin Mining

https://youtu.be/5c2sqhh3IGg?si=Jkxzagg76ZKsFCb7

Check

'So he shows a big box with two oddly placed USB connectors in plain site in a box with apparently with no vents. Even more distinctive is a big off-centred cable entering the front of the box, how many big miners have that arrangement?'

The Block Proto mining rig is still somewhat of a mystery. I haven't been able to find a working model to compare the cable arrangement.

My guess is still that 'the rig' could be a Proto rig that has undergone some design changes since the concept was announced which could be the number and location of power supply reduced, ethernet connection and other modifications post the Proto rig unveiling last year.

Two ethernet connectors could be a design feature; Link aggregation (LACP): If you use two cables with supported hardware (managed switch and NIC with LACP), you can improve throughput and reduce bottlenecks in high-traffic scenarios.

Reminds me of Sphere 3D Corp unveiling a new flashy mining rig that was actually a server some years ago :-)

How do you deal with someone who spent three days staring at a photo of a new mining rig that doesn't have AVAX written on the casing but swears he saw AVAX on the casing ?

USB connectors ? Do you mean Ethernet connectors ?

Coin Dance is a useful website to compare BTC, BCH and BSV. eCash is also shown but as far as I'm aware is only promoted by Jeff Booth ( the really creepy guy) as a 2nd layer payment solution for BTC

https://coin.dance/

A few years ago I suggested QBT could solo mine Bitcoin Cash, the hashrate is far less than the BTC hashrate and it is usually more profitable to mine.

Using any of their optimizations to solo mine BCH blocks could have projected a strong product and it would only have cost a few thousand dollars to buy a rig and start mining. If QBT had mined several BCH blocks regardless of their optimizations the speculation alone could have sent QBT into hundreds of million of pounds.

It might be that QBT now mine Bitcoin Cash during testing.

News of blocks mined could send the SP up and to the right

Red Panda Mining put out this video a month ago:

Solo Mining Bitcoin Cash on My Own Node! (Umbrel OS)

https://youtu.be/MC77p2nCzhs?si=bvcge6g06f-fXkMu

Bitcoin Cash is Up +400% 📈in 3 years and is Up +44% in 12 months 📈

BTC is down -18% 📉in 12 months and flat over 2 years

It's worth noting that BTC isn't the only Bitcoin worth mining

Block are founding members of COPA Crypto Open Patent Alliance so IF the Rig being tested is Blocks Proto rig and the test results are positive I see no reason to dwell on the Patents subject.

COPA members pledge never to use their crypto-technology patents against anyone, except for defensive reasons, effectively making their patents freely available for all to use.

I had the impression QBT had no idea how to market any of their optimizations if they work with all the stories about how they expected to derive revenue, the best thing could simply be that the optimizations work and a big player takes over QBT

If QBT can prove test results that an ASIC manufacturer deems important to the future of their business QBT shares should be as easy to buy as Unobtainium

If test results fail to impress or fail completely that would obviously be a major setback

QBT have progressed with the ASIC UltraBoost Patents which are not cheap, so presumably they must believe or know there's something worth spending 10s or 100s of thousands of pounds on to protect that optimization

DSFlat, and with a selection of losing hashes, what do the stats look like? How many digits 0-9 in the last 45 characters, and the sum?

There were A LOT of losing hashes, too many for me to examine manually lol but as an extreme example a losing hash could presumably have had 45 9's or 45 0's

Are the stats from your selection of winning hashes statistically significant?

They are if miners are brute forcing hashes that are unlikely to be winning hashes...

That's why we're here isn't it ?

QBT have been scanning Bitcoin winning hash data to come up with methods to optimize that data into increased mining performance

Could winning hashes share similar characteristics and can miners and mining pools hone in on some of those winning hashes ?

Hexadecimal is a number system with a base of 16, meaning it uses 16 unique symbols to represent values. These symbols are the digits 0–9 and the letters A–F, where A represents 10, B is 11, C is 12, D is 13, E is 14, and F is 15

000000000000000000010f2d208ddab5c8c46ac11b079cb56cbcf84e53826316 25 out of 45 are the digits 0–9 added together = 111

00000000000000000001d9565a7412e8e6f2190cd3ca8cc1bc4acc5bac037bb3 24 out of 45 are the digits 0–9 added together = 98

000000000000000000010be5f5a971fdc9a046022a50dca25de8e3a793c47c34 27 out of 45 are the digits 0–9 added together = 121

000000000000000000009d07ddb8671bd93711d0631b4b0297b3ae8bdfe29791 26 out of 44 are the digits 0–9 added together = 131

000000000000000000014fb59908fb4dc1697840ced4c6f7d5b86f6e4560fe82 28 out of 45 are the digits 0–9 added together = 142

000000000000000000008e47633843f84dbd0dde19eda90c61e476e271b0eb5e 26 out of 44 are the digits 0–9 added together = 116

00000000000000000001480dbdaaaa2c26ad71b4519dde50fa7b6ca4a4b56e08 23 out of 45 are the digits 0–9 added together = 95

00000000000000000001652e5075b133a0fbe71b4f0d592c8a1e750299b0aaa2 29 out of 45 are the digits 0–9 added together = 109

000000000000000000018cc9ed58316d167772115b692feb42d37b7512912502 33 out of 45 are the digits 0–9 added together = 145

00000000000000000001373f5667be886ac72760b7e3220a6438fc71fad33290 32 out of 45 are the digits 0–9 added together = 140

Having simply looked at some winning hashes over 24 hours I would say yes, firmware targeting set parameters monitored and adjusted by AI could increase winning hashes imo

Will QBT's AI Oracle be first to market and industry leading, we'll have to wait and see

A winning Bitcoin hash is a 64-character hexadecimal number that starts with a specific number of leading zeros, determined by the current network difficulty. It must be less than or equal to the target hash set by the Bitcoin protocol, which adjusts every 2016 blocks to maintain a 10-minute block time.

For example, a valid winning hash might look like:

0000000000000000057fcc708cf0130d95e27c5819203e9f967ac56e4df598ee

So not joking, just compared 17 winning hashes today to see how random there were and if they shared any characteristics

I was quick to dismiss some of QBT's claims in the past regarding optimizations using historical winning hash data that could dismiss a large portion of likely losing hashes without actually looking at historical winning hashes.

I looked at some winning hashes today to see if there could any patterns that might be helpful to an AI Oracle:

Numbers 0 to 9

letters a to f

8 out of 10 winning hashes started with a 1 and were 45 characters long, contain 23 to 33 numbers the sum of which was between 95 and 145

With a possible high number choice of 405 (45X9) and a low number choice of 0 (45X0)

The cumulative number hot zone could contain between 23 and 33 numbers with a cumulative sum total between 95 and 145

Presumably QBT's AI Oracle could focus on hot zones like this, reevaluating hot zones as each block is mined

Example: 00000000000000000001373f5667be886ac72760b7e3220a6438fc71fad33290

Just putting that out there