Quantum Blockchain Technologies (QBT), a research company based in the United Kingdom, has developed artificial intelligence-powered algorithms that could significantly increase the mining winning probability of certain ASIC Bitcoin (BTC) miners, CEO Francesco Gardin said in an interview with Cointelegraph.
Speaking exclusively to the publication, Gardin unpacked how Quantum Blockchain Technologies (QBT) has incorporated AI to enable the smart search of winning hashes as an alternative to conventional random searches.
In the space of two years, the company has developed a number of different patented methods by tapping into the expertise of some twenty experts from the fields of quantum computing, machine learning, cryptography, ASIC chips design and algorithm optimization theory.
QBT’s machine learning teams have developed two different algorithmic search methods which reportedly improve performance of ASIC miners by increasing efficiency and winning result probabilities.
“Method A” is said to improve miner efficiency by 10% while “Method B” is set to improve the probability of a miner finding a winning has by 260%.
Gardin said that the company is looking to explore three specific areas, starting with a short term target of increasing mining performance of existing commercial ASIC chips by adding a software AI component running on a mining rig.
The team is also designing a new architecture for ASIC mining chips to optimize Bitcoin mining, which it detailed in a recent patent application.
Meanwhile QBT has a long term goal of using quantum computers to mine Bitcoin using an in-development SHA-256 computation method that can operate on quantum computing systems.
QBT announced a patent application in July 2023 for the latter, outlining its architectural change to Bitcoin mining ASIC chips which it claims pre-processes data used by future blocks on the Bitcoin blockchain.
Gardin said the QBT Message Scheduling for Cryptographic Hashing ASIC (MSFCA) is able to perform pre-calculations of future BTC blocks before the current block is closed. The “anticipatory resource efficiency algorithm” reduces logic gates of SHA-256 ASIC architecture.
Logic gates are software or hardware devices that carry out logical operations. According to Gardin, MSFCA allows miners to use less logic gates, lowering energy costs and improving efficiency of ASIC mining hardware.
The firm estimates that miners could free up to 8% of logic gates of SHA-256 ASIC chips by pre-processing data used by future blocks on the Bitcoin blockchain, which would make certain logic gates involved in the computation of that data no longer necessary on the ASIC chip.
Gardin also weighed in on the potential for these new methods to influence the Bitcoin mining industry. The QBT said that BTC mining is highly dependent on hardware configurations and hashing power of miners as well as expending considerable amounts of energy.
Gardin added that the probability of finding the winning hash increases with the number and halving speed of a miners’ fleet or an entire pool as well as the corresponding cost of energy while conducting a “completely random search”.
“There are no methods, intelligence, or strategy in current BTC mining, but simply brute force and luck.”
Although the mining rigs market is dominated by just a handful of ASIC manufacturers, Gardin believes that there are minimal differences, features or distinct advances between hardware aside from differences in hashing rates and power consumption.
He added that QBT’s technology, which was mainly developed using Intel’s Blockscale ASIC chips which were recently pulled from production, would provide advantages to any mining rig.
The firm’s technologies are being touted to give an “uncatchable advantage” using AI and SHA-256 optimization and while QBT does not plan to open source its patented methods, Gardin said QBT is considering different options to take its solutions to the Bitcoin mining market.
This could include subscription, licensing, forming a joint venture or outright purchase of the company and its associated technologies.