Firmus Grid

Our first site is under construction in Tasmania, which will support the state's energy grid as it builds its Battery of the Nation ("BotN") project. In addition, we have partnered with leading local technology providers in the design and development of hardware to take advantage of our ultra low PUE advantage. The design and manufacture of custom... PCB and control units directly in Tasmania is proudly fostering the development of a home-grown, critical future industry for Australia. Our data centre is built from the ground up to provide energy-grid-stabilising services, whilst performing as-a-service (AAS) applications (such as machine learning, genome sequencing & research, AI image processing, blockchain proving). Proprietary cooling technology powers a step change in operational efficiency, with a PUE of less than 1.03. Firmus Grid's clean, large load helps transition energy networks towards 100% renewable power. The adoption of renewable energy on a growing scale is great, but it comes with some challenges for the power network that we rely upon. Primary amongst these are consistency of frequency at 50hz (in Australia), and supply matching demand for total power use across the network. Renewables can have challenges when delivering power at a stable 50hz. In addition, they can be liable to sudden changes such as output rising when the sun comes out (in the case of solar), or sudden drop off in the case of wind subsiding. Firmus is able to dial up or down its power usage very rapidly, in response to requests from the network operator (AEMO). This change in our power draw helps to smooth the flow of power through the network. Importantly, it does not cause any loss of data in our computing operations. Working with leaders in cooling systems and fluid engineering, we have developed our own proprietary immersion cooling solution that delivers a world class PUE of 1.03, meaning for every 100kw of computing capacity our centres use less than 3kw of power for all mechanical services. Traditional air cooled data ...Read More