Yttrium Iron Garnet-Based Combinatorial Logic and Memory Devices

Yttrium iron garnet Y3Fe2(FeO4)3 (YIG) has a uniquely low magnetic damping for spin waves, which makes it a perfect material for magnonic devices. Spin waves typically exist in the microwave frequency range, and their wavelength can be decreased to the nanoscale. Their dispersion in YIG waveguides depends on the strength and orientation of the bias magnetic field. It may be possible to exploit YIG waveguides as field-controlled filters and delay lines. In this work, we describe combinatorial logic and memory devices to benefit YIG properties. An act of computation in the combinatorial device is associated with finding a route connecting the input and output ports. We present experimental data demonstrating the pathfinding in the active ring circuit with YIG waveguide. The ability to search in parallel through multiple paths is the most appealing property of combinatorial devices. Potentially, they may compete with quantum computers in functional throughput.

to spin waves to be implemented for data processing. First, spin waves for magnetic bit read-in and read-out [3]. Third, 33 there is a robust and energy-efficient mechanism for spin 34 wave to voltage and vice versa conversion using multiferroic 35 cells [4], [5], [6]. This approach is of great importance for 36 integrating spin-wave devices with conventional electronic 37 components. 38 Fast amplitude damping is the major physical constraint 39 of spin-wave devices based on conducting ferromagnetic 40 materials. For instance, the damping time in permalloy is 41 about of 1 ns at room temperature, which limits the prop-42 agation length to a few tens of micrometers [7]. However, 43 this technical obstacle is overcome in ferrites-in particular, 44 yttrium iron garnet (YIG) [8]. Spin waves in YIG may show 45 relatively large (e.g., up to 1 cm) coherence length even at 46 room temperature. It makes YIG the best material for spin-47 wave-based devices development. There has been a signif-48 icant progress in YIG-based magnonic logic and memory 49 devices prototyping during the past two decades. [9]. The 50 utilization of phase in addition to charge demonstrated great 51 potential in application to NP problems. For instance, prime 52 factorization was accomplished using a spin-wave interfer-53 ometer [10]. The abovementioned works encompass the road 54 toward magnonic devices, which may complement CMOS 55 in the special task data processing. A comprehensive review 56 on the recent advances in spin-wave logic devices can be 57 found in [11]. In this work, we present experimental data on 58 recently proposed combinatorial logic devices in which an act 59 VOLUME The schematics of the active ring circuit are shown in and detection with microantennas can be found elsewhere [8].

77
The group velocity of spin waves propagating in the waveg-

120
The ability to self-adjust to the auto-oscillation frequency 121 is the most appealing property of the active ring circuit.

122
In case condition (3.1) is satisfied, the system is naturally 123 searching for the frequency ω s to satisfy condition (3.2). This 148 is the electron gyromagnetic ratio. k is the wavenumber,  Here, we report experimental data on the spin-wave redi-176 rection in a multiport structure where signal changes its 177 propagation path depending on the position of the external 178 phase shifter . The schematics of the experimental setup 179 are shown in Fig. 3(a). The electric part consists of an ampli-180 fier and a phase shifter. The electric part is connected to a 181 multiport passive path, which consists of a substrate with 182 microantennas covered by the 6-mm × 18-mm YIG film of 183 thickness 21 µm and a saturation magnetization of 1750 G.

184
There are six microantennas made on PCB, which serve 185 as input-output ports. The photograph of the substrate is 186 shown in Fig. 3(b). The antennas are marked as 1, 2, 3, . . . , 6.

187
Antenna #1 is the input port, while the other five antennas are 188 the output ports. An external magnetic field of 270 Oe was 189 applied in the YIG-film plane, as it is shown in Fig. 3(b).

190
At such a direction of the bias magnetic field, spin waves  interesting physical phenomenon that can be used for 237 magnetic bit addressing and read-out.

238
3) There may be a variety of practical applications of 239 the active ring circuits with a multipath passive part 240 using different materials and signal delay mechanisms. 241 Spin-wave delay lines are convenient due to the small 242 size and ability to control dispersion by the bias field. 243 For instance, a voltage-controlled spin-wave modula-244 tor based on the synthetic multiferroic structure was 245 recently demonstrated [6].

246
In Fig. 5, a combinatorial device is schematically shown, 247 which can be utilized for data storage memory and data 248 processing. For simplicity, it shows an active ring circuit 249 with just one amplifier and one phase shifter. The magnetic 250 part is a 5 × 5 matrix of YIG waveguides. The signal can 251 propagate on horizontal and vertical waveguides connecting 252 the nearest-neighbor sites. Each waveguide serves a delay line 253 and a frequency filter. There are n inputs and n outputs. Each 254 input-output port has a switch to control connectivity to the 255 matrix. Also, each output port has a phase shifter to control 256 condition (3.2). There are two positions for a switch: on 257 and off. There are a number of connection combinations, for 258 example, one input-one output combination, two input-one 259 output port, and so on. There are 2 2n−2 possible combinations 260 for the 2n input-output switches. There should be at least 261 one input and one output port connected to the electric part.

289
The overall functional throughput of combinatorial logic 290 devices can be estimated as follows:

291
Functional throughput = 2 2n−2 · (n + n)!/(n! × n!) l 2 · n 2 × l · n 2 /v g (7) 292 where l 2 · n 2 is the size of the multipath matrix, l is the 293 characteristic size of the mesh cell, l · n 2 /v g is the time 294 delay, and v g is the spin-wave group velocity. Regardless 295 of the size of the delay line and signal propagation speed, 296 the functional throughput of the combinatorial logic devices 297 increases proportionally to n factorial. Potentially, combina-298 torial logic devices may compete with quantum computers 299 in functional throughput. The traveling salesman person and 300 the Königsberg bridge problems are NP-hard problems to 301 be solved with the help of combinatorial logic devices. The 302 examples of finding a route through selected points on the 303 mesh and finding the shortest route are described in [12].

304
There are certain physical restrictions on the number of 305 possible paths, which can be recognized. For instance, the 306 number of distinct phases as well as distinct amplitude levels 307 per output is restricted by the accuracy of the phase shifters 308 and attenuators. Also, YIG waveguides have a limited fre-309 quency interval for spin-wave propagation, which implies 310 another technical challenge for engineering magnonic matrix. 311 These and other questions deserve a special consideration. 312 This work is aimed to describe YIG-based combinatorial 313 logic devices and outline their most appealing properties.

387
He is currently a Research Engineer with the 388 University of California, Riverside, Riverside, CA, 389 USA. He has authored or coauthored six book 390 chapters, more than 100 research articles, and 391 seven U.S. patents. His research interests include 392 spin-wave logic devices, magnetometers, multifer-393 roics, electromagnetic energy storage devices, and 394 electromagnetic templates for biomedical applications.