Simon Granville
Simon Granville
Senior Scientist
Principal Investigator of MacDiarmid Institute and a Senior Scientist in the Robinson Research Institute, Victoria University of Wellington, Wellington (New Zealand)
Research Field: Magnetism Without Angular Momentum
Dr. Simon Granville is a Senior Scientist at the Robinson Research Institute of Victoria University of Wellington in New Zealand and a Principal Investigator of the MacDiarmid Institute for Advanced Materials and Nanotechnology. He is an experimental materials physicist working on magnetic materials and devices, especially thin film rare earth nitrides and Heusler alloys. His current projects include developing energy-efficient, ultra-high-speed and high-performance magnetic memory and logic for superconducting computing; spintronic sources of THz waves; and magnetic sensors for monitoring infrastructure. He also investigates the spintronic and spin caloritronic properties of ferromagnetic Weyl semi-metal thin films and devices.
Magnetic semiconductor rare earth nitrides – spintronic materials for energy-efficient cryogenic computing memories
S. Granville 1,2, K. van Koughnet 2,3, E. Trewick 2,3, C. Pot 2,3, W. Holmes-Hewett 2,3, J. D. Miller 2,3,
R. G. Buckley 1,3, H. J. Trodahl 2, and B. J Ruck 2,3
1 Robinson Research Institute, Victoria University of Wellington, New Zealand
2 School of Chemical and Physical Sciences, Victoria University of Wellington, New Zealand
3 MacDiarmid Institute for Advanced Materials and Nanotechnology, New Zealand
At current trends, energy use for computing will reach 40% of the world’s electricity production and 10% of all the world’s energy production, by the early 2030s [1,2]. Even if existing transistor-based processing continues to improve as the Si technology base approaches long-touted physical limits, it will soon be imperative to change to a less energy-intensive form of computing.
Cryogenic computing based on superconductors could provide a solution for replacing the most energy-intensive data centres and cloud computing facilities [3]. However, there is no type of memory operable at cryogenic temperatures that can make superconducting computing realistic. Spintronic memory types have the scalability, energy-efficiency and non-volatility required, but the metallic magnetic materials used in them do not meet the needs of hybrid magnetic-superconducting memories.
The rare earth nitrides (RENs) are series of magnetic semiconductors well-suited for superconducting electronics – they have highly tuneable conductivities, saturation magnetisations and coercive fields spanning orders of magnitude [4,5]. They can also be alloyed to achieve zero net magnetisation at full magnetic alignment [6], ideal for high-speed memories in superconducting environments without fringe magnetic fields. In this talk I will cover our studies of thin films and structures made from RENs and (RE1, RE2)N alloys, towards solving the problem of memory for scalable superconducting logic, as needed in energy-efficient cryogenic computer concepts, such as quantum computers. I will conclude by describing our concept cryogenic MRAM structures that are can be read by and integrated directly with superconducting electronics.
Reference:
[1] Decadal Plan for Semiconductors, Semiconductor Research Corporation (2021)
https://www.src.org/about/decadal-plan/
[2] D. Natelson, The need for energy-efficient computing [blog post] (2022, November 15)
[3] D. S. Holmes et al., Computer 48, 34 (2015).
[4] C.-M. Lee et al., Appl. Phys. Lett. 106, 022401 (2015)
[5] A. Shaib et al., AIP Adv. 11, 015125 (2021)
[6] J. D. Miller et al., Phys. Rev. B 106, 174432 (2022)