Quantum Blockchain Technologies (AIM:QBT), the AIM-listed investment company focused on blockchain research and development, said its team has completed the reconfiguration of its Method C AI Oracle software to run on an ASIC manufacturer's mining rig.
The adaptation was more complex than expected, requiring the underlying learning models to be reconfigured rather than simply retrained, because the rig's rolling architecture differs materially from the Bitaxe Gamma system on which the Oracle was originally developed.
Subsequent testing showed a consistent AI Oracle advantage over traditional mining across the evaluation windows completed so far.
QBT presented its progress to the ASIC manufacturer on 28 July, confirming that an initial AI model is now running as a baseline for further refinement.
A follow-up session with the manufacturer is scheduled for the coming weeks, when additional data should allow further performance gains toward the levels already achieved on the Bitaxe Gamma.
Separately, the company has been compressing the Oracle for direct deployment on the mining rig's own hardware, with early tests of a compressed version already showing a significant improvement in processing speed.
"We are pleased with the progress made in adapting our AI models to the ASIC manufacturer's mining rig," said Francesco Gardin, chief executive and executive chairman.
He added that confirming the long-term robustness of results remains crucial before the software can be considered commercially viable.
The next milestone is the live demonstration to the ASIC manufacturer, once testing on its Mining Development Kit is complete.
News Intelligence what this means for the company
QBT has successfully adapted its Method C AI Oracle software to run on an ASIC manufacturer's mining rig after reconfiguring underlying learning models to account for hardware differences, with early testing showing consistent performance gains over traditional mining. The company presented a working baseline model to the manufacturer on 28 July and has scheduled follow-up sessions to refine performance further. This represents incremental progress toward commercial viability, but the CEO emphasized that long-term robustness confirmation remains essential before the software can be considered ready for market.
The adaptation milestone advances QBT's path from R&D to potential commercialization, but the company remains pre-revenue and the software has not yet met the robustness threshold management deems necessary for deployment. Success depends on whether performance gains can be sustained and replicated at scale with the ASIC manufacturer's hardware.
Insights assembled by AI. Editor-reviewed and grounded in tickstock’s coverage and proprietary knowledge graph.
Content is for informational purposes only, not financial advice.
