Periodic Reporting for period 4 - TOPSPIN (Topotronic multi-dimensional spin Hall nano-oscillator networks)
Période du rapport: 2024-03-01 au 2025-08-31
TOPSPIN’s overall aim is to i) fabricate, ii) optimize, iii) functionalize, iv) utilize, and v) explore very large-scale one-, two-, and three-dimensional spin Hall nano-oscillator (SHNO) networks. These nano-oscillators operate at microwave frequencies and can be utilized both for wireless communication and novel types of computing.
The project is divided into three work packages with separate goals as outlined below:
WP1 – Multidimensional arrays, ultra-low currents, and efficient (non-volatile) tuning.
Fabricate.
• Push the capability and quality of e-beam nano-lithography fabrication of SHNOs to demonstrate breakthroughs in multidimensional mutual synchronization: SHNO chains, very large two-dimensional (2D) SHNO arrays, and eventually 3D SHNO arrays.
• Develop nano-lithography processes to fabricate mutually synchronized magnetic tunnel junction based SHNO arrays to combine the best possible SHNO signal coherence with the highest microwave output power.
Optimize.
• Demonstrate SHNOs operating at orders of magnitude lower power consumption through the use of ultra-high spin Hall angle materials.
• Demonstrate low-power SHNOs operating at an order of magnitude higher microwave frequency.
• Demonstrate ultra-high Γp values, using perpendicular and mixed anisotropy bilayers, to increase SHNO synchronization speeds by orders of magnitude.
Functionalize.
• Demonstrate ultra-fast SHNO frequency tuning and tuning of the SHNO interconnection strengths and phases through the use of voltage-controlled magnetic anisotropy.
• Demonstrate non-volatile SHNO and interconnection tuning using ion oxide materials and phase change materials.
WP2 – Microwave signal generators, pattern matching, and neuromorphic computing.
Utilize.
• Deliver the first viable 10 – 300 GHz spintronic microwave signal generators.
• Demonstrate ultra-fast and ultra-efficient pattern matching using SHNO chains.
• Demonstrate highly scalable neuromorphic computing using synchronized SHNO chains and arrays.
WP3 – Novel characterization techniques for unrivaled time-, phase-, and spatial resolution.
Explore.
• Time- and phase-resolved Brillouin Light Scattering microscopy of mutually synchronized large-scale SHNO chains and arrays.
• Laser tuning and excitation of individual SHNOs in large arrays while observing both the non-local and global network response.
• Scanning transmission x-ray microscopy of SHNOs for extreme spatial resolution.
• Ultra-fast Lorentz microscopy of SHNOs for extreme time- and spatial resolution.
The project is divided into three work packages with separate goals as outlined below:
WP1 – Multidimensional arrays, ultra-low currents, and efficient (non-volatile) tuning.
Fabricate.
• Push the capability and quality of e-beam nano-lithography fabrication of SHNOs to demonstrate breakthroughs in multidimensional mutual synchronization: SHNO chains, very large two-dimensional (2D) SHNO arrays, and eventually 3D SHNO arrays.
• Develop nano-lithography processes to fabricate mutually synchronized magnetic tunnel junction based SHNO arrays to combine the best possible SHNO signal coherence with the highest microwave output power.
Optimize.
• Demonstrate SHNOs operating at orders of magnitude lower power consumption through the use of ultra-high spin Hall angle materials.
• Demonstrate low-power SHNOs operating at an order of magnitude higher microwave frequency.
• Demonstrate ultra-high Γp values, using perpendicular and mixed anisotropy bilayers, to increase SHNO synchronization speeds by orders of magnitude.
Functionalize.
• Demonstrate ultra-fast SHNO frequency tuning and tuning of the SHNO interconnection strengths and phases through the use of voltage-controlled magnetic anisotropy.
• Demonstrate non-volatile SHNO and interconnection tuning using ion oxide materials and phase change materials.
WP2 – Microwave signal generators, pattern matching, and neuromorphic computing.
Utilize.
• Deliver the first viable 10 – 300 GHz spintronic microwave signal generators.
• Demonstrate ultra-fast and ultra-efficient pattern matching using SHNO chains.
• Demonstrate highly scalable neuromorphic computing using synchronized SHNO chains and arrays.
WP3 – Novel characterization techniques for unrivaled time-, phase-, and spatial resolution.
Explore.
• Time- and phase-resolved Brillouin Light Scattering microscopy of mutually synchronized large-scale SHNO chains and arrays.
• Laser tuning and excitation of individual SHNOs in large arrays while observing both the non-local and global network response.
• Scanning transmission x-ray microscopy of SHNOs for extreme spatial resolution.
• Ultra-fast Lorentz microscopy of SHNOs for extreme time- and spatial resolution.
The TOPSPIN project has achieved its goals and delivered major advances in nanomagnetism, spintronics, and unconventional computing. Over its full duration, the project produced 39 peer-reviewed publications and an additional preprint under review, with many results appearing in leading scientific journals. All three work packages progressed strongly and led to several breakthroughs in device fabrication, oscillator physics, computing concepts, and ultrafast characterization.
A central achievement was the creation of large, coherent networks of spin Hall nano-oscillators (SHNOs). TOPSPIN demonstrated two-dimensional arrays containing more than 100,000 mutually synchronized oscillators—an improvement of five orders of magnitude compared with the state of the art before the project began. These results establish SHNOs as a powerful platform for studying collective dynamics and for building next-generation computing hardware based on coupled oscillators.
The project also showed how SHNO technology can be combined with memristive elements to create reconfigurable oscillator networks. By integrating non-volatile gates directly on top of SHNOs, we demonstrated selective programming and long-term storage of their oscillation properties, enabling ultrafast pattern-recognition tasks in one-dimensional chains. In parallel, we established two-dimensional SHNO arrays as viable oscillator-based Ising Machines, capable of representing and solving optimization problems in a hardware-native manner.
Another major outcome of TOPSPIN is the development of two advanced microscopy techniques that provide new capabilities for visualizing spin waves, phonons, and oscillator states. The first combines femtosecond laser–comb excitation with Brillouin light scattering microscopy, enabling studies of ultrafast magnetization and phonon dynamics with simultaneously high temporal, spectral, and spatial resolution. The second is a phase-sensitive optical microscope for observing injection-locked and phase-binarized SHNOs, allowing the mapping of their phase states with sub-diffraction-limited precision.
Together, these scientific and technological achievements greatly expand what is experimentally possible in nanoscale magnetism and oscillator-based computing. The results of TOPSPIN establish a strong foundation for future research on coherent nanomagnetic networks, beyond-CMOS hardware, and ultrafast magnetization dynamics.
A central achievement was the creation of large, coherent networks of spin Hall nano-oscillators (SHNOs). TOPSPIN demonstrated two-dimensional arrays containing more than 100,000 mutually synchronized oscillators—an improvement of five orders of magnitude compared with the state of the art before the project began. These results establish SHNOs as a powerful platform for studying collective dynamics and for building next-generation computing hardware based on coupled oscillators.
The project also showed how SHNO technology can be combined with memristive elements to create reconfigurable oscillator networks. By integrating non-volatile gates directly on top of SHNOs, we demonstrated selective programming and long-term storage of their oscillation properties, enabling ultrafast pattern-recognition tasks in one-dimensional chains. In parallel, we established two-dimensional SHNO arrays as viable oscillator-based Ising Machines, capable of representing and solving optimization problems in a hardware-native manner.
Another major outcome of TOPSPIN is the development of two advanced microscopy techniques that provide new capabilities for visualizing spin waves, phonons, and oscillator states. The first combines femtosecond laser–comb excitation with Brillouin light scattering microscopy, enabling studies of ultrafast magnetization and phonon dynamics with simultaneously high temporal, spectral, and spatial resolution. The second is a phase-sensitive optical microscope for observing injection-locked and phase-binarized SHNOs, allowing the mapping of their phase states with sub-diffraction-limited precision.
Together, these scientific and technological achievements greatly expand what is experimentally possible in nanoscale magnetism and oscillator-based computing. The results of TOPSPIN establish a strong foundation for future research on coherent nanomagnetic networks, beyond-CMOS hardware, and ultrafast magnetization dynamics.
TOPSPIN has delivered clear advances beyond the state of the art across fabrication, device physics, computing concepts, and experimental methodology. The project achieved the first demonstration of two-dimensional networks containing more than 100,000 mutually synchronized spin Hall nano-oscillators—an increase of five orders of magnitude compared to the best results available before the project began. This establishes a new performance benchmark for coherent nanoscale oscillator networks and opens the door to large-scale collective-dynamics and computing applications.
The project also achieved the first integration of memristive functionality directly into SHNO devices. This breakthrough enables non-volatile and locally programmable control of individual oscillators and their coupling within networks, providing a new mechanism for reconfigurable, energy-efficient hardware. Building on these capabilities, TOPSPIN further demonstrated that two-dimensional SHNO arrays can serve as hardware Ising Machines, showing for the first time that such spintronic networks can perform optimization tasks in a scalable, oscillator-based architecture.
In addition to device and computing advances, TOPSPIN developed new experimental techniques that significantly extend the state of the art in spin-wave and phonon imaging. We built the first microscope that combines femtosecond laser–comb excitation with Brillouin light scattering, enabling the study of ultrafast spin-wave and phonon dynamics with unprecedented temporal and spectral resolution. A complementary phase-sensitive optical microscope for mapping injection-locked and phase-binarized SHNOs provides sub-diffraction spatial resolution of oscillator phase states, offering unique insight into the behaviour of dense oscillator networks.
These achievements collectively define major progress beyond the state of the art. Although the funded period has concluded, we have a substantial body of unpublished results that we will continue to analyze and develop into future publications. Moreover, the project has equipped our research group with unique, state-of-the-art fabrication, measurement, and microscopy capabilities that are not available elsewhere. These tools will continue to generate high-impact scientific results and drive further advances in spintronics, ultrafast magnetism, and oscillator-based computing well beyond the end of the project.
The project also achieved the first integration of memristive functionality directly into SHNO devices. This breakthrough enables non-volatile and locally programmable control of individual oscillators and their coupling within networks, providing a new mechanism for reconfigurable, energy-efficient hardware. Building on these capabilities, TOPSPIN further demonstrated that two-dimensional SHNO arrays can serve as hardware Ising Machines, showing for the first time that such spintronic networks can perform optimization tasks in a scalable, oscillator-based architecture.
In addition to device and computing advances, TOPSPIN developed new experimental techniques that significantly extend the state of the art in spin-wave and phonon imaging. We built the first microscope that combines femtosecond laser–comb excitation with Brillouin light scattering, enabling the study of ultrafast spin-wave and phonon dynamics with unprecedented temporal and spectral resolution. A complementary phase-sensitive optical microscope for mapping injection-locked and phase-binarized SHNOs provides sub-diffraction spatial resolution of oscillator phase states, offering unique insight into the behaviour of dense oscillator networks.
These achievements collectively define major progress beyond the state of the art. Although the funded period has concluded, we have a substantial body of unpublished results that we will continue to analyze and develop into future publications. Moreover, the project has equipped our research group with unique, state-of-the-art fabrication, measurement, and microscopy capabilities that are not available elsewhere. These tools will continue to generate high-impact scientific results and drive further advances in spintronics, ultrafast magnetism, and oscillator-based computing well beyond the end of the project.