Welcome to Atlas
Our team has developed patented IP for microchip designs that optimise the storage and operation of Graph Databases, an upcoming technology which unlocks previously unattainable insight and decision-making capabilities for digital businesses.
About Atlas
We are a research group from the University of Edinburgh specialising in AI hardware innovations.
Our team want to explore the commercialisation of a plug-and-play hardware accelerator for graph databases (GDBs). Unlike current databases that store data in cumbersome tables, GDBs encode the connections and relationships between data in an easily navigable structure, promising not only order-of-magnitude improvements in database response speed and power dissipation but also enabling new AI paradigms (see “hybrid AI”).
However, there is still enormous untapped potential: all market-leading implementations of GDBs rely on general parallel computing. We have designed a chip prototype that realises these improvements in speed and power.
Atlas Graph Benefits
Our chip design intrinsically stores data as graphs, meaning that data points that are connected via relationships in a database are also connected in memory. It is specifically developed to support core GDB operations and search algorithms, thereby eliminating operating inefficiencies that normally stem from applying generic hardware to highly specific software. This allows denser storage which also offers increases in query speed by up to 100x and decreases in energy consumption by up to a half. Our IP is highly parameterizable and can seamlessly be integrated into existing plug-and-play hardware solutions, designed to power the next generation of AI applications. Such hardware will be of significant value to major Cloud service providers (AWS, Azure, etc.) which suffer from scaling, energy consumption and bandwidth challenges in an age where data usage is increasing exponentially.
Graph Databases (GDBs) are an emerging technology which stores data and encodes the corresponding relationships between them. This powerful architecture allows businesses to identify previously hidden relationships between their data and the value offered to their customers. However, the increase in the amount of data stored worldwide and the need to closer to real-time processing are seriously challenging Cloud service providers who rely on conventional computing technologies to handle GDBs at scale. Our technology rethinks GDB computing in a bottom-up approach which aims to make GDB storage denser, reduce their associated carbon footprint and dramatically increase their query speeds.
Faster query speeds
%
Reduction in energy costs
Meet the Atlas team
Dr Chris Giotis
Early Career Researcher
Dr Christos Giotis Christos Giotis is a Research Associate in the Centre for Electronics Frontiers at the University of Edinburgh. He holds a PhD in Electronic Engineering from the same group where he developed hardware synapses for continuous online learning. His expertise lies in the intersection between software and hardware and he is currently leading the efforts behind commercialising the Atlas technology. He is currently conducting market exploration to identify how Atlas can best add value to key segments in the Graph Database market.
Dr Alex Serb
Senior Researcher
Dr Alex Serb is a Reader in Unconventional AI Hardware Technologies at the University of Edinburgh. He is a MIET and SMIEEE. He has been a key architect and leading member of the delivery team of £20M+ worth of projects. He acts as the main technical coordinator of the effort. This involves supervising the technical team working on the technology, working with the applicant to ensure that technical specifications required by commercial imperative are met by the technical team and planning the MVP.
Dougie Thoms
Early Career Researcher
Dougie Thoms is Product Director at Wood Mackenzie. He is leading development efforts to consolidate commodity market datasets (supply, demand, trade, prices) across the industries Wood Mackenzie covers to provide customers with the confidence they need to make better investment decisions through the Energy Transition. He serves as the Atlas’ external Business Advisor providing guidance throughout the team’s market exploration efforts and access to his relevant network.
Veronica Ferguson
Early Career Researcher
Veronica Ferguson is an experienced project manager with a passion for people, enterprise and entrepreneurs. She has been responsible for supporting the direct investment of £750,000 in grant funding and £3m in equity investment into 6 companies over the past 12 months and personally supported 2 companies to spin out of the University in the academic year 2021-2022. She is Staff Enterprise Team Manager for the University of Edinburgh’s academic staff working with high-growth potential projects from idea/concept through to growth and investment stages. She is based in Edinburgh Innovations and works closely with the Technology Transfer Manager with responsibility for IP as well as other key colleagues to support the commercialisation process.
Our Partners
FAQs
You probably have lots of questions about the Atlas Graph technology and how it can be applied in your business so we’ve answered some of our frequently asked questions here. If you can’t see an answer to your question, then please get in touch with our team!
Will Atlas Graph work for me?
Your content goes here. Edit or remove this text inline or in the module Content settings. You can also style every aspect of this content in the module Design settings and even apply custom CSS to this text in the module Advanced settings.
Will Atlas Graph work for me?
Your content goes here. Edit or remove this text inline or in the module Content settings. You can also style every aspect of this content in the module Design settings and even apply custom CSS to this text in the module Advanced settings.
Will Atlas Graph work for me?
Your content goes here. Edit or remove this text inline or in the module Content settings. You can also style every aspect of this content in the module Design settings and even apply custom CSS to this text in the module Advanced settings.
Contact Us
Get in touch with the Atlas Graph team to learn more about our project and ask any questions you may have about our graph database technology. We’d love to hear from you!
© 2023 Atlas Graph | University of Edinburgh