JP7687569B2 - SEU cross section estimation device, SEU cross section estimation method, and SEU cross section estimation program - Google Patents

Description

The present invention relates to an SEU cross-section estimation device, an SEU cross-section estimation method, and an SEU cross-section estimation program.

When cosmic rays raining down from space collide with oxygen or nitrogen in the atmosphere, neutron rays are generated and rain down on the Earth. On the other hand, when neutron rays enter circuits such as LSIs used in many electronic devices, a soft error occurs in which the bits of data stored in the LSI are flipped due to the electric charge generated by the nuclear reaction.

Soft errors are known as SEUs (Single Event Upsets), and occur when a single particle of radiation (neutrons, protons, heavy particles, etc.) enters a semiconductor device such as a memory, causing the data (bits) stored in the semiconductor device to be inverted due to charges generated by a nuclear reaction. The percentage of particle radiation that causes SEUs is called the SEU cross-section.

If the number of SEUs when a semiconductor is irradiated with particle radiation of fluence φ [n/cm2] is N, the SEU cross section is given by the following equation (1).

(SEU cross section)=N/φ (1)
Known methods for measuring SEU cross sections include the time-of-flight method disclosed in, for example, Non-Patent Documents 1 and 2 and Patent Document 1. The time-of-flight method is a method in which the speed of particle radiation is calculated by measuring the time it takes to fly a certain distance, and the speed is converted into particle energy.

By generating pulsed neutrons using an accelerator and a nuclear reactor, placing a detector at a certain distance, and measuring the difference between the time the pulsed neutrons are generated and the time the neutrons are detected, i.e., the time of flight, it is possible to identify the energy of the neutrons.

https://group.ntt/jp/newsrelease/2020/11/25/201125a.html Hidenori Iwashita et al. “Energy-Resolved Soft-Error Rate Measurements for 1-800 MeV Neutronsby the Time-of-Flight Technique at LANSCE” IEEE TRANSACTIONS ON NUCLEAR SCIENCE, VOL. 67, NO. 11, NOVEMBER2020.

International Publication No. 2021/002469

However, to measure high-energy neutrons on the order of MeV, a large-scale facility is required, and there are only a limited number of facilities worldwide that can perform such measurements. It is difficult to measure SEU cross-sections using such a large-scale facility. Furthermore, if the device to be measured is an FPGA (Field Programmable Gate Array), a high-speed detection circuit can be incorporated to measure the speed of neutrons using the time-of-flight method, but for devices such as SRAM in which it is difficult to incorporate a high-speed detection circuit, soft errors cannot be detected at high speed. For example, there is a problem in that scanning the entire memory takes a long time, from several milliseconds to tens of milliseconds.

The present invention has been made to solve the above problems, and has as its object to provide an SEU cross-section estimation device, an SEU cross-section estimation method, and an SEU cross-section estimation program that are capable of estimating the SEU cross-section of a semiconductor device whose SEU cross-section is unknown in a simple manner.

The SEU cross-section estimation device of one aspect of the present invention includes a memory unit that stores the soft error occurrence rate for each of a plurality of reference devices having known SEU cross-sections when a plurality of targets made of different materials are irradiated with each neutron generated by irradiating each of the targets with particle radiation, a calculation unit that calculates the soft error occurrence rate when a measurement target device is irradiated with each of the neutrons, a determination unit that determines the similarity between the soft error occurrence rate of the measurement target device and the soft error occurrence rate of the reference device, and an estimation unit that estimates the SEU cross-section of the measurement target device based on the SEU cross-section of the reference device that is determined to have the highest similarity to the measurement target device.

The SEU cross-section estimation method of one aspect of the present invention includes the steps of: storing the soft error occurrence rate for each of a plurality of reference devices, whose SEU cross-sections are known, when a plurality of targets made of different materials are irradiated with neutrons generated by irradiating each of the targets with particle radiation; calculating the soft error occurrence rate when a measurement target device is irradiated with the neutrons; determining the similarity between the soft error occurrence rate of the measurement target device and the soft error occurrence rate of the reference device; and estimating the SEU cross-section of the measurement target device based on the SEU cross-section of the reference device determined to be most similar to the measurement target device.

One aspect of the present invention is an SEU cross-section estimation program for causing a computer to function as the above-mentioned SEU cross-section estimation device.

The present invention makes it possible to estimate the SEU cross section of a semiconductor device whose SEU cross section is unknown in a simple manner.

FIG. 1 is a block diagram showing the configuration of an SEU cross-section estimation device according to an embodiment and its peripheral devices. FIG. 2 is an explanatory diagram showing how a target is irradiated with protons to generate neutrons. FIG. 3 is a graph showing the relationship between neutron energy and the number of neutrons generated when protons are irradiated onto a number of targets made of different materials. FIG. 4 is a graph showing the relationship between neutron energy and SEU cross section when a number of existing devices are irradiated with neutrons. FIG. 5 is a graph showing the soft error occurrence rate when neutrons generated using targets made of a variety of materials are irradiated onto each reference device and a measurement target device. FIG. 6 is a flowchart showing a processing procedure performed by the SEU cross-section estimation device according to the first embodiment. FIG. 7 is an explanatory diagram showing how a target having a layered structure made of a first material and a second material is irradiated with protons to generate neutrons. FIG. 8 is a graph showing the relationship between neutron energy and the number of neutrons generated when protons are irradiated onto a plurality of targets having different thicknesses of the second material. FIG. 9 is a graph showing the soft error occurrence rate when each reference device and each measurement target device are irradiated with neutrons generated using a plurality of targets having different thicknesses of the second material. FIG. 10 is a block diagram showing the hardware configuration of this embodiment.

[Description of the First Embodiment]
The SEU cross-section estimation apparatus according to the embodiment will be described below with reference to the drawings. Fig. 1 is a block diagram showing the configuration of the soft error estimation apparatus according to the embodiment (hereinafter, abbreviated as "estimation apparatus 1") and its peripheral devices, that is, an accelerator 2, a target 3, and a measurement target device 4.

The accelerator 2 accelerates and outputs particle radiation such as protons and heavy particles. In this embodiment, an example in which protons are used as particle radiation will be described.

When the target 3 is irradiated with protons output from the accelerator 2 as shown in FIG. 2, it outputs neutrons. The target 3 is made of a material such as beryllium (Be), lead (Pb), tantalum (Ta), or tungsten (W), and is installed in an appropriate position on the output side of the accelerator 2. By changing the material of the target 3, the neutron fluence (amount of neutrons emitted per unit area) and neutron spectrum emitted from the target 3 change.

The neutron spectrum is data showing the relationship between neutron energy and the number of neutrons. The neutron fluence and neutron spectrum for each material of the target 3 are known data. Note that in this embodiment, examples are described in which beryllium (Be), lead (Pb), tantalum (Ta), and tungsten (W) are used as the material of the target 3, but the present invention is not limited to this, and targets 3 made of other materials may also be used.

Figure 3 is a graph showing the neutron spectrum emitted from target 3, with the horizontal axis showing neutron energy [MeV] and the vertical axis showing the number of neutrons. Curves p1, p2, p3, and p4 shown in Figure 3 show the neutron spectrum emitted from target 3 when the material of target 3 is tantalum (Ta), beryllium (Be), tungsten (W), and lead (Pb), respectively. The neutron fluence and neutron spectrum data are output to the estimation device 1.

The measurement target device 4 is a semiconductor device for which the SEU cross section is to be estimated. In other words, the SEU cross section of the measurement target device 4 is unknown. The measurement target device 4 is a board on which semiconductors such as FPGAs and SRAMs (Static Random Access Memory) are mounted. When the measurement target device 4 is irradiated with neutrons output from the target 3, a soft error occurs in the measurement target device 4. Information on the soft error that occurs in the measurement target device 4 is output to the estimation device 1.

The estimation device 1 (SEU cross-section estimation device) includes an input unit 11, a calculation unit 12, a calculation data storage unit 13, an existing data storage unit 14, a memory unit 15, a determination unit 16, an estimation unit 17, and an output unit 18.

The input unit 11 acquires information on the neutron fluence irradiated from the target 3 when the target 3 is irradiated with protons from the accelerator 2. The input unit 11 acquires information on the occurrence of soft errors that occur when the measurement target device 4 is irradiated with neutrons. When the material of the target 3 is changed to various materials, the input unit 11 acquires information on the neutron fluence for each target 3 of each material, and information on soft errors that have occurred in the measurement target device 4.

Specifically, the input unit 11 acquires neutron fluence information and soft error information when the target 3 is made of beryllium (Be), lead (Pb), tantalum (Ta), and tungsten (W). The input unit 11 outputs the acquired information to the calculation unit 12.

The calculation unit 12 calculates the soft error occurrence rate for each target (i.e., for each neutron spectrum) based on the information of the soft errors that occurred in the measurement target device 4. Specifically, the calculation unit 12 calculates the soft error occurrence rate for each target 3 based on the information of the soft errors that occurred in the measurement target device 4 when the targets 3 are beryllium (Be), lead (Pb), tantalum (Ta), and tungsten (W). The calculation unit 12 calculates the soft error occurrence rate, for example, as shown by symbols P1 to P4 in FIG. 5, which will be described later. That is, the calculation unit 12 calculates the soft error occurrence rate when the measurement target device 4, which is the subject of measurement of the SEU cross-section, is irradiated with neutrons generated by irradiating each target 3 with particle radiation.

The calculation data storage unit 13 stores the soft error occurrence rate calculated by the calculation unit 12.

The existing data storage unit 14 stores data on the SEU cross sections of multiple FPGAs. The SEU cross sections of each FPGA are known data that have been measured in advance using well-known techniques such as time-of-flight and monochromatic methods. In the following, a semiconductor device such as an FPGA whose SEU cross section is known is referred to as a "reference device." In this embodiment, an example will be described in which "FPGA 28 nm," "FPGA 40 nm," and "FPGA 55 nm" are used as reference devices. The existing data storage unit 14 stores data on the SEU cross sections for each of "FPGA 28 nm," "FPGA 40 nm," and "FPGA 55 nm."

Note that "28 nm", "40 nm", and "55 nm" refer to the line width formed on the FPGA substrate. For example, "FPGA 28 nm" is an FPGA on which a pattern with a line width of 28 nm is formed. In this embodiment, an example is described in which an FPGA is used as the reference device, but the present invention is not limited to this, and other semiconductor devices may be used.

Figure 4 is a graph showing the SEU cross-sections of each reference device, with the horizontal axis showing neutron energy [MeV] and the vertical axis showing SEU cross-section [cm2/Mbit]. Curve p11 in Figure 4 shows the SEU cross-section of a 28 nm FPGA, curve p12 shows the SEU cross-section of a 40 nm FPGA, and curve p13 shows the SEU cross-section of a 55 nm FPGA.

The memory unit 15 stores the soft error occurrence rate for each reference device due to neutrons generated using various targets 3. That is, it stores the soft error occurrence rate for each different neutron spectrum. As shown in FIG. 3, the soft error occurrence rate for each reference device due to neutrons generated using various targets 3 is calculated based on a graph showing the relationship between neutron energy and the number of neutrons when targets 3 made of multiple materials are used, and the SEU cross-section data of multiple reference devices (FPGAs) shown in FIG. 4.

Specifically, as shown in FIG. 5, data is obtained with the horizontal axis representing the target 3 (Be, Pb, Ta, W) and the vertical axis representing the soft error occurrence rate, and this data is stored in the memory unit 15. FIG. 5 shows the soft error occurrence rate of each reference device normalized by the soft error occurrence rate of a 28 nm FPGA. Curve p21 in FIG. 5 shows the soft error occurrence rate of a 28 nm FPGA, curve p22 shows the soft error occurrence rate of a 40 nm FPGA, and curve p23 shows the soft error occurrence rate of a 55 nm FPGA. That is, the memory unit 15 stores the soft error occurrence rate for each target 3 when multiple reference devices with known SEU cross sections are irradiated with neutrons generated by irradiating multiple targets 3 made of different materials with particle radiation.

The determination unit 16 compares the soft error occurrence rate of the measurement target device 4 stored in the calculation data storage unit 13 with the soft error occurrence rate of the reference device stored in the memory unit 15, and determines which reference device the soft error occurrence rate of the measurement target device 4 is similar to. In other words, the determination unit 16 determines the similarity between the soft error occurrence rate of the measurement target device 4 and the soft error occurrence rate of the reference device.

For example, assume that the soft error occurrence rates occurring in the measurement target device 4 when neutrons are irradiated to the measurement target device 4 using targets 3 made of various materials are symbols P1, P2, P3, and P4 as shown in FIG. 5. In this case, it is determined that there is a high similarity between the soft error occurrence rate of the measurement target device 4 and the soft error occurrence rate of the FPGA 55 nm shown in curve p23. One specific example of a determination method is a determination method using the least squares method. Each soft error occurrence rate is aP1, aP2, aP3, and aP4, and the coefficient a that minimizes the sum of squares of the residuals of each of the model curves p23, p22, and p11 is searched for, and the model curve with the smallest sum of squares of the residuals is determined to be the similar curve.

When the determination unit 16 identifies a highly similar reference device, the estimation unit 17 recognizes the relationship between the soft error occurrence rate of this reference device and the soft error occurrence rate of the measurement target device 4. For example, as shown in FIG. 5, the estimation unit 17 recognizes that the soft error occurrence rate of the measurement target device 4 is similar to the soft error occurrence rate of the FPGA 55 nm, and furthermore, that the soft error occurrence rate is about 20% lower than that of the FPGA 55 nm.

The estimation unit 17 estimates that the SEU cross section of the device 4 to be measured is about 20% lower than the SEU cross section of the 55 nm FPGA shown in the curve p23. The estimation unit 17 estimates that the SEU cross section of the device 4 to be measured is about 20% lower than the SEU cross section of the 55 nm FPGA shown in the curve p13 of FIG. 4.

That is, the estimation unit 17 calculates the ratio between the soft error occurrence rate of the reference device determined by the determination unit 16 to be highly similar and the soft error occurrence rate of the measurement target device 4, and estimates the SEU cross section of the measurement target device 4 based on the SEU cross section of the reference device determined to be highly similar and the ratio.

The output unit 18 outputs the data estimated by the estimation unit 17 to the outside.

Next, the procedure for estimating an SEU cross section in the estimation device 1 shown in the first embodiment will be described with reference to the flowchart shown in FIG.

First, in step S11 shown in FIG. 6, the existing data storage unit 14 acquires and stores data showing the relationship between neutron energy and SEU cross-sections of multiple reference devices whose SEU cross-sections are known. Specifically, data on the SEU cross-sections of the FPGA 28 nm, FPGA 40 nm, and FPGA 55 nm shown in FIG. 4 is stored. The cross-sections of the reference devices are obtained, for example, by the time-of-flight method using a large-scale facility.

In step S12, the input unit 11 acquires the soft error occurrence rate when the measurement target device 4 is irradiated with neutrons generated by irradiating multiple targets 3 with protons. Specifically, data for points P1 to P4 shown in FIG. 5 is acquired.

In step S13, the determination unit 16 identifies a reference device with a high degree of similarity in soft error occurrence rate to the measurement target device 4. Specifically, since the data at points P1 to P4 shown in FIG. 5 changes similarly to the curve p23, the highly similar reference device is identified as an FPGA 55 nm.

In step S14, the estimation unit 17 recognizes the relationship between the soft error occurrence rate of the measurement target device 4 and the soft error occurrence rate of the reference device identified in the processing of step S13. Specifically, it recognizes that the data of points P1 to P4 shown in FIG. 5 is about 20% lower than the graph of the FPGA 55 nm. Furthermore, it estimates the SEU cross section of the measurement target device 4 based on the SEU cross section of the FPGA 55 nm.

Specifically, it is estimated that the SEU cross section of the device 4 to be measured will be approximately 20% lower than the curve p13 (SEU cross section of a 55 nm FPGA) shown in FIG. 4. This estimation result is output from the output unit 18. Then, this process ends. Based on this estimation result, the user can recognize the SEU cross section of the device 4 to be measured.

As described above, the estimation device 1 (SEU cross-section estimation device) according to this embodiment includes a memory unit 15 that stores the soft error occurrence rate for each of multiple targets 3 when multiple reference devices with known SEU cross-sections are irradiated with neutrons generated by irradiating multiple targets 3 made of different materials with particle radiation, a calculation unit 12 that calculates the soft error occurrence rate when the measurement target device 4 is irradiated with each neutron, a determination unit 16 that determines the similarity between the soft error occurrence rate of the measurement target device 4 and the soft error occurrence rate of the reference device, and an estimation unit 15 that estimates the SEU cross-section of the measurement target device 4 based on the SEU cross-section of the reference device that is determined to have the highest similarity to the measurement target device 4.

By using the estimation device 1 according to this embodiment, the SEU cross section of a semiconductor device (device 4 to be measured) whose SEU cross section is unknown can be estimated using the SEU cross section of a reference device. This makes it possible to estimate the SEU cross section of device 4 to be measured in an extremely simple manner without using large-scale facilities.

In addition, beryllium (Be), lead (Pb), tantalum (Ta), and tungsten (W) are used as targets 3. By irradiating targets 3 of each material with protons, different neutron spectra can be generated, making it possible to estimate the SEU cross-section with high accuracy.

The determination unit 16 also calculates the ratio between the soft error occurrence rate of the reference device determined to have a high similarity and the soft error occurrence rate of the measurement target device 4. For example, in the example shown in FIG. 5, the soft error occurrence rate (P1 to P4) of the measurement target device 4 is determined to be approximately 20% lower than the soft error occurrence rate (curve p23) of the FPGA 55 nm. The estimation unit 17 estimates the SEU cross section of the measurement target device 4 based on this ratio, thereby improving the estimation accuracy of the SEU cross section.

In addition, it becomes possible to easily estimate the SEU cross section even if the device being measured is one in which it is difficult to incorporate a high-speed detection circuit, such as an SRAM.

[Description of the Second Embodiment]
Next, a second embodiment will be described. The estimation apparatus 1 according to the second embodiment has a similar configuration to that of the first embodiment. In the second embodiment, the structure of the target 3 that generates neutrons to be irradiated to the measurement target device 4 is different from that of the first embodiment.

The second embodiment also differs from the first embodiment in that the storage unit 15 stores a graph showing the relationship between the thickness of aluminum (second material 32) and the soft error occurrence rate, as shown in FIG. 9, which will be described later. The second embodiment will be described in detail below.

Figure 7 is an explanatory diagram showing the configuration of a target 30 used when estimating an SEU cross-section in the estimation device 1 according to the second embodiment. As shown in Figure 7, the estimation device 1 according to the second embodiment uses a target 30 with a two-layer structure in which a first material 31 is laminated with a second material 32.

As an example, a target 30 is used that is formed by stacking a first material 31 with a constant thickness (e.g., 3 mm) and a second material 32 with a variable thickness. The first material 31 is, for example, beryllium (Be), and the second material is, for example, aluminum (Al). By using multiple targets 30 with different aluminum thicknesses, the neutron spectrum output from the target 30 is changed and irradiated onto the device 4 to be measured.

In the first embodiment described above, multiple targets 3 made of different materials are used, whereas in the second embodiment, multiple targets 30 made of different thicknesses of the second material 32 are used to generate neutrons. That is, the multiple targets 30 include a first target and a second target, and the first target and the second target have a layered structure of a member made of the first material (first material 31) and a member made of the second material (second material 32), and the member made of the second material of the first target and the member made of the second material of the second target have different plate thicknesses. The memory unit 15 stores the soft error occurrence rate for each target 30 when each reference device is irradiated with neutrons generated by irradiating the first target and the second target with protons (particle radiation).

Figure 8 is a graph showing the neutron spectrum when the thickness of the aluminum is changed in the range of 0.0 to 1.5 mm. Curves q1, q2, q3, q4, and q5 shown in Figure 8 are graphs showing the neutron spectrum when the thickness of the aluminum used as the second material 32 is 0.0 mm, 0.2 mm, 0.5, 1.0 mm, and 1.5 mm, respectively.

The memory unit 15 stores a graph showing the relationship between the aluminum thickness and the soft error occurrence rate for each reference device, as shown in FIG. 9. The graph shown in FIG. 9 shows data normalized by the soft error occurrence rate of a 28 nm FPGA. The curve q21 shown in FIG. 9 shows the soft error occurrence rate for a 28 nm FPGA, the curve q22 shows the soft error occurrence rate for a 40 nm FPGA, and the curve q23 shows the soft error occurrence rate for a 55 nm FPGA.

Points Q1, Q2, Q3, and Q4 in FIG. 9 indicate the soft error occurrence rate in the measurement target device 4 when the measurement target device 4 is irradiated with neutrons using multiple targets 30 with different aluminum thicknesses.

That is, in the second embodiment, the memory unit 15 stores the soft error occurrence rate for each target 30 when neutrons generated by irradiating particle radiation onto multiple targets 30, each of which is made of a first material with a fixed thickness and a second material with various thicknesses, are irradiated onto each reference device.

As can be seen from FIG. 9, points Q1 to Q4 are similar to curve q22. Therefore, the judgment unit 16 judges that the soft error occurrence rate of the measurement target device 4 is highly similar to the FPGA 40 nm reference device.

In addition, points Q1 to Q4 have a soft error occurrence rate that is about 20% higher than that of curve q22. The estimation unit 17 estimates that the SEU cross section of the measurement target device 4 will be a value that is 20% higher than that of the SEU cross section of the 40 nm FPGA. Specifically, it estimates that the value will be 20% higher than that of curve p12 shown in Figure 4.

In this way, in the second embodiment, neutrons are generated using a target 30 having a layered structure of a first material 31 (a member made of a first material) and a second material 32 (a member made of a second material) with the thickness of the second material 32 being varied, and the SEU cross section of the device 4 to be measured is estimated based on the soft error occurrence rate detected using each target 30 having a different thickness of the second material 32.

Therefore, in the estimation device 1 according to the second embodiment, as in the first embodiment described above, the SEU cross section of a semiconductor device (device 4 to be measured) whose SEU cross section is unknown can be estimated using the SEU cross section of a reference device. This makes it possible to estimate the SEU cross section of device 4 to be measured without using large-scale facilities.

In the estimation device 1 according to the second embodiment, neutrons are generated using multiple types of targets 30 in which the thickness of the second material 32 is appropriately changed. This eliminates the need to prepare targets of many materials, making it possible to further simplify the estimation of the SEU cross-section.

The SEU cross-section estimation device (estimation device 1) according to the embodiment described above may be, for example, a general-purpose computer system including a CPU (Central Processing Unit, processor) 901, a memory 902, a storage 903 (HDD: Hard Disk Drive, SSD: Solid State Drive), a communication device 904, an input device 905, and an output device 906, as shown in FIG. 10. The memory 902 and the storage 903 are storage devices. In this computer system, the CPU 901 executes a predetermined program loaded on the memory 902, thereby realizing each function of the estimation device 1.

The estimation device 1 may be implemented in one computer, or in multiple computers. The estimation device 1 may also be a virtual machine implemented in a computer.

The program for the estimation device 1 can be stored on a computer-readable recording medium such as a HDD, SSD, USB (Universal Serial Bus) memory, CD (Compact Disc), or DVD (Digital Versatile Disc), or can be distributed via a network.

The present invention is not limited to the above embodiment, and many variations are possible within the scope of the invention.

1. Estimation device (SEU cross-section estimation device)
2 Accelerator 3, 30 Target 4 Measurement target device 11 Input section 12 Calculation section 13 Calculation data storage section 14 Existing data storage section 15 Memory section 16 Determination section 17 Estimation section 18 Output section 31 First material (member of first quality)
32 Second material (member made of second material)

Claims (8)

a memory unit that stores a soft error occurrence rate for each target when a plurality of reference devices having known SEU cross sections are irradiated with neutrons generated by irradiating a plurality of targets made of different materials with particle radiation;
a calculation unit that calculates a soft error occurrence rate when the device to be measured is irradiated with each of the neutrons;
a determination unit that determines a degree of similarity between a soft error occurrence rate of the measurement target device and a soft error occurrence rate of the reference device;
an estimation unit that estimates an SEU cross section of the device under test based on an SEU cross section of a reference device that is determined to have the highest similarity to the device under test;
An SEU cross-section estimation apparatus comprising: the estimation unit calculates a ratio of a soft error occurrence rate of the reference device determined by the determination unit to have the highest similarity to a soft error occurrence rate of the measurement target device;
The SEU cross-section estimation apparatus according to claim 1 , further comprising: estimating an SEU cross-section of the measurement target device based on the ratio and the SEU cross-section of the reference device determined to have the highest similarity. 3. The SEU cross-section estimation device according to claim 1, wherein the material of each of the targets is any one of beryllium (Be), tantalum (Ta), lead (Pb), and tungsten (W). the plurality of targets include a first target and a second target, the first target and the second target have a laminated structure of a member made of a first material and a member made of a second material, and the member made of the second material of the first target and the member made of the second material of the second target have different plate thicknesses;
3. The SEU cross-section estimation device according to claim 1, wherein the memory unit stores a soft error occurrence rate for each target when neutrons generated by irradiating the first target and the second target with the particle radiation are irradiated to each of the reference devices. The SEU cross-section estimation device according to claim 4 , wherein the first material member is beryllium (Be) and the second material member is aluminum (Al). 6. The SEU cross-section estimation device according to claim 1, wherein each of the targets generates a different neutron spectrum when irradiated with the particle radiation. a step of storing a soft error occurrence rate for each target when a plurality of reference devices having known SEU cross sections are irradiated with neutrons generated by irradiating a plurality of targets made of different materials with particle radiation;
calculating a soft error occurrence rate when the device to be measured is irradiated with each of the neutrons;
determining a degree of similarity between a soft error occurrence rate of the measurement target device and a soft error occurrence rate of the reference device;
estimating an SEU cross-section of the device under test based on an SEU cross-section of a reference device determined to have the highest similarity to the device under test;
The SEU cross-section estimation method includes: An SEU cross-section estimation program that causes a computer to function as the SEU cross-section estimation device according to any one of claims 1 to 6.

Images