Our Portfolio and Current Assets

Quantum Transportation

The Critical Role of Error Correction

Error correction, a critical element of quantum control, emerged as a key innovation in 2024, with Google’s Willow quantum computing chip demonstrating significant advancements in error correction and performance. With 105 physical qubits, Willow can perform certain complex calculations exponentially faster than supercomputers and with a low error rate.Start-ups, too, showed progress on error correction. For example, Alice & Bob presented a new quantum error correction architecture, Riverlane unveiled a hardware-based quantum error decoder with enhanced speed and efficiency, QuEra launched a logical quantum processor based on reconfigurable atom arrays,and Atom Computing collaborated with Microsoft to deliver quantum error correction.

Ensuring QT systems are less prone to error is essential for achieving the stability and accuracy needed to deploy quantum applications at scale.

The Problem

Noisy Qubits Limit Quantum Computing Qubits — the fundamental units of quantum computers — are inherently noisy. Without quantum error correction (QEC), large-scale quantum computation is impossible. Efficient QEC requires both a well-designed code and a powerful decoder — the algorithm that interprets erroneous data to identify and correct errors. Decoding remains a key bottleneck. Addressing it would significantly improve noise tolerance, enhancing the performance and scalability of today’s quantum hardware. For hardware manufacturers, this is a game-changer.

Since 1998, when QEC was proven to be a possible reality, until 2020, the focus of the hardware companies was the Surface code (SC), a specific QEC code. This code is relatively easy to implement and analyze for hardwares but it is very wasteful- it needs several physical qubits to encode one unit of meaningful information. A useful, efficient fault tolerant quantum computer will not be based on the Surface code. And justifiably so, the landscape now is rapidly changing.

Solution In Detail

Solution

A Patented Machine Learning Based Universal Decoder Unlike traditional decoders, the patented decoder (PD) is a machine learning based decoder that estimates and refines errors using advanced neural network techniques. The decoder is Code-agnostic, i.e., it generalizes naturally to any stabilizer or CSS code (surface, color, bicycle, product codes, and beyond). It is noise-aware, i.e., it adapts to the actual noise model, training directly on realistic channel data. It is Scalable, i.e., once trained, it can be applied across different hardware platforms and code sizes. All of this makes PD uniquely scalable, adaptable, and essential for future quantum hardware.

In Detail- Product description

Our PD and future our services/ assistance provides a platform for Quantum hardware companies and/ or labs to research and choose an error correction scheme that best suits their hardware needs. Our focus will be to assist small- medium sized companies & labs that cannot afford to have a QEC research team in-house. This simulator will also facilitate in-house research that would yield original academic papers & further patents. To that end, the plan is to upgrade the simulator to expand and include Code Design.