Rahko is solving chemistry with quantum machine learning 

 

Rahko specialises in quantum machine learning for faster and more accurate simulation of drugs and materials for the discovery and development of new molecules and materials at greatly reduced cost. 

Fundamental areas of research and technology are critically limited by: high cost, low speed and low accuracy. 

Rahko is using quantum machine learning to remove these limitations and unlock radical advances. 

The Rahko customer journey starts with solutions that can be deployed today for far greater speed at far lower cost, and progresses to readiness to exploit the high precision that quantum computers will offer as soon as they are at scale for powerful practical applications. 

 

Faster + cheaper today 

Rahko is developing the world’s most advanced quantum machine learning models and software to enable our customers to accurately predict chemical properties at a much higher speed and lower cost than traditional methods (with speedups up to 5 orders of magnitude). 

These models and software can make immediate use of novel simulation methods and tools. 

They are also able to integrate with quantum computers, which will allow them to be used to deliver higher precision once quantum computers have matured for powerful practical applications. 

 

Faster + cheaper + highest precision tomorrow 

Classical computers, even the world’s largest supercomputers, are limited in computational power. This limits the accuracy with which we can simulate the behaviour of drugs and materials. 

Low accuracy in simulation is one of the key barriers to our ability to discover and develop new drugs, batteries, advanced materials and chemicals. 

Quantum computers will offer far greater computational power than classical computers. Current quantum computers, however, are not yet at the scale needed for powerful practical applications.

The quantum machine learning models Rahko is able to deploy today will be able to be integrated into quantum computers when the devices are at scale. 

 

Quantum machine learning – more than just a necessary complement to quantum hardware 

Near-term quantum computers (‘noisy-intermediate-scale quantum’ or ‘NISQ’ computers) are highly prone to error, commonly termed ‘noise’. 

Quantum machine learning algorithms have been proven to be remarkably resilient to noise by Rahko and a small number of teams across the world. 

Rahko focuses on 2nd generation quantum machine learning algorithms, ‘parametrized quantum circuits’ that can be trained like deep learning models and can be highly robust to noise on near-term quantum computers. 

This will allow us to ensure our customers are ideally positioned to fully exploit the first advantages that will be offered by quantum computers of scale.