Hello everyone,
My name is Ahmad Shajahan, a 2nd-year B.Tech Computer Science student at APJ Abdul Kalam Technological University, Kerala, India.
I have recently completed an original research paper titled:
“Thermodynamic Neural Computation (TNC): Novel Free-Energy Minimisation with Cache-Aware Memory-Mapped Annealing for Energy-Efficient AI.”
The paper proposes a new computational framework where gradient descent is replaced by thermodynamic free-energy minimisation, introducing 14 novel equations aimed at improving energy efficiency in AI training.
Key idea:
Using thermodynamic principles and cache-aware annealing mechanisms to achieve significantly more energy-efficient optimization compared to standard SGD.
The paper is publicly available on Zenodo:
https://doi.org/10.5281/zenodo.19029046
ORCID:
https://orcid.org/0009-0001-3287-3055
I am preparing to submit this work to arXiv (cs.LG – Machine Learning), but as a first-time submitter I require an endorsement from an existing arXiv author.
If anyone eligible to endorse in cs.LG is willing to review or endorse the submission, I would be very grateful.
I would be happy to share the full manuscript or discuss the work.
Thank you very much for your time and support.
— Ahmad Shajahan