LOOPUS. MATHEMATICAL PROBLEM SOLVING CIRCUIT INCLUDING RESISTIVE ELEMENTS
Introduction
LOOPUS is a breakthrough in computing systems, enabling scalable, zero-latency, energy-efficient solution to machine learning and big data analytics via analogue in-memory accelerators. In particular it develops analogue accelerators for algebraic computing, delivering your machine learning and big data analytics in just one click.

Technical features
Loopus is an innovative hardware accelerator for machine learning (ML) and big data processing. A new electronic circuit has been developed to train ML algorithms, including linear / logistic regressions, neural networks and page ranking in one step. The circuit is based on non-volatile analog memories and feedback systems (hence the name Loopus).
The implementation of the ML algorithms is divided into two phases: training and inference. The first is to send a known database to a network, teaching it, for example, how to classify inputs. This is realized with time-consuming algorithms able to optimize the network parameters for a given problem.
The concept of loop computing has been developed to accelerate the training phase for cloud and edge computing, thus saving time and costs for data centers and allowing low processing power for artificial intelligence (AI).
Possible Applications
- Big data analysis;
- Machine learning.
Advantages
- In-memory analogue computation, where data analysis is performed directly in analogue memory with no need to transfer data;
- Physical matrix-vector moltiplication within a XP circuit thanks to Kirchoff’s Law and Ohm’s Law;
- Physical iterqation thanks to the feedback loop instead of time-consuming numerical iteration.