Universal Memory
for the AI Era
DRAM-Class Speed
Non-Volatile Storage
Ultra-Low Energy

Awards

ULTRARAMâ„¢

Tech Overview

ULTRARAMâ„¢ is a remarkable, patented, memory technology developed by Lancaster University. It exploits a quantum-mechanical process called resonant tunnelling, allowing ULTRARAMâ„¢ to deliver non-volatility with fast and energy-efficient write and erase, resulting in high endurance. This combination of properties was thought to be unachievable until now. ULTRARAMâ„¢ has the non-volatility of flash, with a performance that is expected to exceed that of DRAM.
The $165bn pa memory market is dominated by dynamic random-access memory (DRAM, $100bn) and NAND flash ($60bn). Flash is non-volatile, retaining data when unpowered, but is slow and has poor program/erase cycling endurance. In contrast, DRAM is fast with excellent endurance, but it is volatile, and requires data to be constantly refreshed. For decades there has been a quest for a memory that combines their advantages without their disadvantages, i.e. a memory that is fast and non-volatile, with high endurance and ultra-low switching energies; a so-called ‘universal memory’. ULTRARAM™ has achieved these universal memory characteristics.

Explainer Video

AI is Limited by Memory

ULTRARAMâ„¢

Artificial intelligence is transforming every industry, but the energy cost of moving data is becoming one of the biggest constraints on future computing performance.

From hyperscale AI data centres to intelligent edge devices, memory is increasingly the bottleneck.

AI is Limited by Memory

ULTRARAMâ„¢

ULTRARAMâ„¢

Today’s Memory Hierarchy Was Not Built for AI

  • Data movement dominates system energy consumption
  • Memory bandwidth limits performance
  • DRAM and flash require constant trade-offs between speed, power and persistence

The future of AI requires a different approach.

A New Class of Memory

ULTRARAMâ„¢ combines the speed of DRAM with the persistence of flash memory in a single device.

Built using patented quantum resonant tunnelling structures within III-V semiconductors, ULTRARAM offers a fundamentally new approach to storing and accessing data.

Potential Advantages

  • High-speed access
  • Non-volatile operation
  • Ultra-low energy consumption
  • Long endurance
  • Simplified memory architectures

Program Erase / Switching Energies

10,000 x

Lower than other emerging memories

100 x

Lower than DRAM

Program Erase / Switching Energies

1,000 x

Lower than Flash

Enabling the Future of AI Infrastructure

By reducing the energy and performance penalties associated with data movement, ULTRARAM has the potential to support the next generation of:

AI Data Centres:

Reducing memory power consumption and improving system efficiency.

Edge AI:

Enabling intelligent devices to operate closer to where data is created.

High-Performance Computing:

Supporting data-intensive applications that demand both speed and efficiency.

Secure & Sovereign Computing:

Providing new opportunities for defence, critical infrastructure and secure systems.

Introducing ULTRARAM-Neuroâ„¢

Extending the ULTRARAM platform into compute-in-memory and neuromorphic architectures.

ULTRARAM-Neuroâ„¢ is being developed to explore new approaches to AI acceleration by bringing memory and computation closer together.

Future Applications

  • Compute-in-memory AI

  • Agentic edge AI

  • Neuromorphic systems

  • Ultra-low power inference

  • Autonomous intelligent platforms

ULTRARAMâ„¢

Proven Device Physics. Global Ecosystem.

ULTRARAMâ„¢ is based on patented technology developed at Lancaster University and is supported by an international ecosystem of semiconductor partners and research collaborators.
Collaborations
Protected by a growing international patent portfolio.

ULTRARAMâ„¢

Building a Platform for the Next Generation of Computing

Quinas is developing a new memory platform designed for the demands of artificial intelligence, high-performance computing and future intelligent systems.
Our mission is simple:
Eliminate the memory bottleneck limiting the future of AI.

Partner With Us

Quinas is working with investors, industrial partners, foundries, researchers and system developers to bring ULTRARAMâ„¢ to market.

Our Team

QuInAs Team
James Ashforth-Pook
Chief Executive Officer
Manus Hayne
Chief Scientific Officer
Peter Hodgson
Chief Technology Officer

Our Advisors

QuInAs Advisors
Prof Avirup Dasgupta
Dr. J. Iwan Davies

Contact Us

UK HQ

QUINAS Technology Limited
167-169 Great Portland Street 5th Floor
London W1W 5PF
United Kingdom

Registered in England and Wales.
Company Registration No. 14673840.

R&D Labs

QUINAS Technology
A035, Faraday Buildings
Physics Avenue, Bailrigg
Lancaster LA1 4YB
United Kingdom

EU Design Center

QuInAs Tech (Malta) Ltd
Sir Temi Zammit Buildings
San Gwann Industrial Estate
San Gwann, SGN 3000
Malta

Quinas Tech (Malta) Ltd
Ewropa Business Centre,
Level 3, Suite 701,
Dun Karm Street,
B’ Kara, BKR9034, Malta

APAC

APAC

TAIWAN TECH ARENA
No.2, Sec. 4, Nanjing E. Rd., Songshan Dist.,
Taipei City 105037, Taiwan (R.O.C.)
Taiwanarena.tech | Quinas

Send us a message

Privacy Policy

We are committed to protecting and preserving your privacy. Your contact details and email address will not be shared outside the QuInAs team. See our full Privacy Policy.

Investors Contact

You are an interested investor ?

Acknowledgements

ULTRARAM is developed in partnership with Lancaster University and funded by the EPSRC (under IAA grant EP/X525583/1 and the Future Compound Semiconductor Manufacturing Hub grant EP/P006973/1), by the Materials Social Futures doctoral training Program of the Leverhulme Trust, by the European Commission via ATTRACT (grants 777222 and 101004462), by MSCA-ITN QUANTIMONY (grant 956548) and by the Innovate UK ICURe program.

MEMBERS

Members