GB/T 47468-2026Performance testing methods for quantum computing systems (English PDF)
量子计算系统性能测试方法
This is a limited preview
Buy now to download the full PDF (31 pages)
Need quotation for your management?
Get a formal quotation for GB/T 47468-2026 in 30 minutes.
Issued by
SAMR; SAC
Level / Type
National · Recommended
Issue date
April 30, 2026
Implementation date
November 1, 2026
Scope
GB/T 47468-2026 is the English-translated version of 量子计算系统性能测试方法.
GB/T 47468-2026 describes the performance metrics and the test methods for quantum computing systems.
The metrics are grouped into device specifications - total, controllable and readable qubit numbers, connectivity, qubit survival rate and the coherence times T1 and T2 - basic control indicators such as initialization, single-qubit gate, two-qubit gate and readout fidelity, the gate and readout durations, quantum gate capacity, manipulation parallelism and calibration capability, and comprehensive indicators including the maximum number of entangled qubits, fidelity from random circuit sampling and idle and operating energy consumption.
Its test procedures include Ramsey and CPMG sequences for transverse relaxation. Clause 1 states that it applies to performance testing during design, development, manufacturing and use, and not to non-gate quantum computing systems. Issued on 30 April 2026, it applies from 1 November 2026.
Found what you need?
Document preview — GB/T 47468-2026
National Standard of the People's Republic of China
- ICS
- 35.240.01
- Classification
- L 70
Issued by: State Administration for Market Regulation; Standardization Administration of the PRC
Contents
- 1 Scope
- 4 Abbreviations
- 6 Testing Methods
- 7 Test Report
1 Scope
This document describes the performance metrics and testing methods for quantum computing systems.
This document applies to performance testing of quantum computing systems during design, research and development, manufacturing, and use. This document does not apply to non-gate quantum computing systems.
4 Abbreviations
The following abbreviations apply to this document.
CLOPS. Number of quantum circuit layers that can be executed per second. CNOT. Controlled NOT CPMG. Pulse sequence (Carr-Purcel-Meiboom-Gil)
CZ. Controlled Z MQC. Many-body Quantum Coherence
POVM. Positive Operator-Valued Measure
QV. Quantum Volume
XEB. Cross-Entropy Benchmarking
c) Number of readable qubits. Characterizes the number of qubits whose quantum state information can be measured with high fidelity when the readout fidelity reaches 85% or higher. Number of bits.
Note: Quantum bits are divided into data qubits and auxiliary qubits. Data qubits are qubits used to encode quantum information, while auxiliary qubits are... Specifically, it refers to qubits added to assist data qubits in completing computational tasks.
5.2.2 Connectivity
Connectivity includes the following metrics.
a) Quantum bit connectivity. characterizes the number of qubits that are directly entangled by a single qubit.
b) Connectivity of a single quantum processor. the average value representing the connectivity of all qubits.
Note: This document does not consider the connectivity of multiple quantum processors.
5.2.3 Bit Survival Rate
The qubit survival rate characterizes the ratio of the number of qubits that are functioning normally to the total number of qubits in a quantum processor.
5.2.4 Bit coherence
The coherence of a quantum bit includes the following metrics.
a) Longitudinal relaxation time T1.Characterizes the time required for a qubit to spontaneously relax from an excited state (high energy state) to the ground state (low energy state).
The degree reflects the rate at which energy is dissipated between the quantum bit and its environment, also known as the energy relaxation time.
b) Lateral relaxation time T2.Characterizes the time scale by which a qubit maintains its quantum state phase coherence, reflecting the phase phase of the qubit.
The rate at which dryness decays over time, also known as the phase relaxation time, includes the effect of energy relaxation.
c) Pure phase relaxation time Tphi. Characterizes the pure phase of the qubit's state after removing the influence of energy relaxation from the transverse relaxation.
Relaxation characteristic time. Using the measured values of T1 and T2, the relationship between Tphi and T1 and T2 can be obtained using formula (1).
5.3 Basic Control Indicators
5.3.1 Fidelity
Fidelity includes the following metrics.
a) Initialization fidelity. Characterizes the relationship between the actual operational effect and the ideal expected effect after initializing the qubit. Similar to a measure of degree.
b) Single-bit gate fidelity. Characterizes the actual operational performance versus the theoretical expected performance after performing a single-bit gate operation on a single qubit. A measure of the similarity between results.
c) Two-qubit gate fidelity. Characterizes the actual operational effect versus the theoretical expected effect after performing a two-qubit gate operation on two qubits. A measure of the degree of similarity between them.
d) Readout fidelity. Characterizes the probability that the measurement result of a qubit matches the true state of the qubit, and determines the qubit's position. The accuracy of the measurement.
5.3.2 Duration
Duration includes the following metrics.
a) Bit initialization time. Characterizes the time required for all readable qubits to be initialized from an unknown state when the initialization fidelity reaches 95% or higher. The time required to reach the initial state.
b) Single-bit gate duration. Characterizes the time required to execute a single-bit gate operation when the single-bit gate fidelity reaches 98% or higher.
c) Two-bit gate duration. Characterizes the time required to execute two-bit gate manipulation when the fidelity of the two-bit gate reaches 95% or higher.
d) Reading time. Characterizes the time required to complete a single high-fidelity measurement of a quantum state when the read fidelity reaches 85% or higher.
5.3.3 Quantum gate capacity
The quantum gate capacity characterizes the number of quantum gate operations that can be completed in coherent time. A larger value indicates a greater number of operations that can be completed in coherent time.
The more sub-operations there are, the better the balance between the coherence and operational efficiency of the qubit.
5.3.4 Manipulating Parallelism
The degree of parallelism in manipulation includes the following metrics.
a) Parallelism of single-qubit gate manipulation. Characterizes the maximum number of simultaneous operations that a quantum computing system can perform under a controllable number of qubits.
If single-qubit gate manipulation is performed on N physical qubits and the fidelity exceeds 98%, then the single-qubit gate of this quantum computing system is... The parallelism of the operation is N.
b) Parallelism of two-qubit gate manipulation. Characterizes the maximum number of simultaneous operations that a quantum computing system can perform under a controllable number of qubits.
N groups (two physical qubits per group) perform two-qubit gate manipulation on two physical qubits, with fidelity exceeding 95% in each group.
The parallelism of the two-bit gate manipulation in this quantum computing system is N.
c) Parallelism of readout manipulation. Characterizes the degree of parallelism of reading and manipulation within the limit of the number of readable qubits in a quantum computing system, where at most N physical qubits can be read simultaneously.
If the readout manipulation on a rational qubit achieves a fidelity exceeding 85%, then the parallelism of readout manipulation in this quantum computing system is N.
5.3.5 System calibration capability
System calibration capability includes the following indicators.
a) The system has an automatic calibration function. Calibration includes single-bit gate parameters, two-bit gate parameters, parameter readings, and quantum calibration. The fidelity of the gate and the read.
b) Calibration cycle. The period from when the characterization system completes the last calibration until the fidelity index drops below the threshold, at which point the next calibration is initiated. Time interval.
c) Calibration duration. Characterizes the time required for a complete calibration process, including the time from initiating the calibration command to completing the calibration and being ready to perform the task. Total time spent on tasks.
5.4 Comprehensive Performance Indicators
5.4.1 Maximum number of entangled bits
The maximum number of entangled qubits represents the maximum number of direct product states (i.e., non-entangled states) in a many-body quantum system that cannot be decomposed into smaller subsystems.
The number of qubits contained in a large entangled subset reflects the globality and indivisibility of entanglement between qubits in a many-body quantum system.
GHZ states are maximally entangled Schrödinger's cat states. Testing the entanglement fidelity of GHZ states with different qubit numbers can be used to evaluate quantum computing systems. An important reference for system performance.
5.4.2 Number of qubits in the algorithm
Number of qubits in the algorithm ( 5.4.6 Fidelity Based on Random Line Sampling Fidelity based on random circuit sampling refers to the measurement of the difference between a practical quantum processor and an ideal quantum processor by using the output samples of random quantum circuits.
The similarity index between noisy quantum computing systems is one of the key benchmarks for evaluating the performance of current medium-scale noisy quantum processors.
It is a key method for verifying the performance of quantum computing systems and demonstrating quantum superiority.
5.4.7 Energy Consumption
Energy consumption includes the following indicators.
a) Idle power consumption. All subsystems involved in the computational power of the quantum computing system are powered on and running stably, but not in operation. The electrical energy consumed during the computing power test.
b) Operating Energy Consumption. The electrical energy consumed by the quantum computing system when running standard computing power test programs under normal operating conditions, such as random access. Line sampling test procedure.
6 Testing Methods
6.1 Device Specifications
f) Fit the data [t, P1(t)] using formula (6).
6.1.1 Number of qubits
The method for testing the number of qubits includes the following steps.
a) Method for testing the total number of qubits. Obtain the claimed total number of qubits from the system design scheme and test report.
b) Methods for testing the controllable number of qubits. This involves initial state preparation and quantum gate operation fidelity testing (see 6.2.1), ergodicity... The number of controllable qubits is defined as the number of qubits that meet the fidelity threshold requirement and is recorded in the system test report.
c) Method for testing the number of readable qubits. By using a readability fidelity test method, the number of qubits is traversed, and those that meet the fidelity threshold are... The required number of bits is denoted as the number of readable qubits and recorded in the system test report.
6.1.2 Connectivity
The connectivity testing method includes the following steps.
a) Disable qubit mapping and circuit optimization on the compiler system.
b) For the i-th qubit, perform the system native two-qubit ratio between the i-th qubit and the j-th (j!=i) qubit. Special operation.
c) Run the circuit. If there are no alarms or errors during the circuit operation, and the fidelity of the circuit output Fij is greater than the set threshold... If 95% of the edges are connected, then edge Lij is considered valid, meaning qubits i and j are connected. The number of valid edges connected to qubit i is... The connectivity K(Gi) of this qubit.
d) Calculate connectivity. Take the average connectivity of all qubits using formula (2) as the overall average connectivity of the system.
6.1.3 Bit Survival Rate
The bit survival rate test method includes the following steps.
a) Obtain the total number of qubits N given in the system design scheme and test report;
b) A measurement is performed on a single qubit if and only if the single-qubit gate fidelity is not less than 98% and the two-qubit gate fidelity is not less than 98%.
If the readout fidelity is not less than 85%, the qubit is considered to be a normally functioning qubit.
c) Perform the measurement process in step b) on all qubits, and count the number of qubits that ultimately satisfy all the above fidelity thresholds.
The quantity is denoted as N1, which is the number of normally functioning qubits.
d) Calculate the bit survival rate according to formula (3);
6.1.4 Bit coherence
6.1.4.1 Longitudinal relaxation time T1
The timing sequence for the longitudinal relaxation time test is shown in Figure 2, and the longitudinal relaxation time is denoted as T1.
Indexing symbol explanation. t --- the free evolution time of a quantum bit.
The longitudinal relaxation time T1 test method includes the following steps.
a) Disable the qubit mapping function on the compilation system and specify the qubit to be measured.
b) Initialize the sub-bit to be measured to the |0 > state.
c) Apply an X gate to flip the sub-bit to be measured to the |1 > state.
d) After the qubit has been free-evolved for time t, the quantum state is measured to obtain the probability P1 that the qubit is in the |1 > state.
e) Gradually extend the free evolution time t to obtain P1 corresponding to different evolution times t, thereby obtaining a set of measurement data [t, P1(t)].
f) Fit the measurement data [t, P1(t)] using formula (4).
6.1.4.2 Lateral relaxation time T2
6.1.4.2.1 Ramsey Experimental Test Method
The timing sequence of the Ramsey experiment for transverse relaxation time is shown in Figure 3, with the transverse relaxation time symbol labeled as T*2.
a) Disable the qubit mapping function on the compilation system and specify the qubit to be measured.
b) Initialize the sub-bit to be measured to the |0 > state.
c) Apply an X/2 operation to the sub-bit to be measured to prepare it into a superposition state at the equator of the Bloch sphere.
d) After the quantum bit has freely evolved for time t, an X/2 operation is applied to the quantum state, and then the quantum state is measured to obtain the quantum bit. The probability P1 of being in the state |1 >.
e) Gradually extend the evolution time t to obtain the probability P1 at different evolution time t, thereby obtaining a set of measurement data [t, P1(t)].
f) Fit the data [t, P1(t)] using formula (5).
6.1.4.2.2 CPMG Experimental Test Method
The timing sequence of the CPMG experimental test method for transverse relaxation time is shown in Figure 4, and the symbol for transverse relaxation time is marked as TCPMG2.
The CPMG test method for transverse relaxation time TCPMG2 includes the following steps.
a) Disable the qubit mapping function on the compilation system and specify the qubit to be measured.
b) Initialize the sub-bit to be measured to the 0 > state.
c) Apply an X/2 operation to the sub-bit to be measured to prepare it into a superposition state at the equator of the Bloch sphere.
d) The phase error is accumulated during the idle time t/2n, and then n Ry(pi) gates are applied sequentially to shift the bits to the other end of the xy plane.
The time interval between Ry(pi) gates is t/n. After the last Ry(pi) gate finishes, wait for t/2n time before applying an Rx(pi/2) gate.
The gatekeeper excites the bit to the 1 > state, and finally measures the probability P1 of the bit being in the 1 > state.
Note 1.n is the number of pi gates applied.
e) Gradually extend the evolution time t to obtain the probability P1 at different evolution time t, thereby obtaining a set of measurement data [t, P1(t)].
6.2 Basic Control Indicators
6.2.1 Fidelity
6.2.1.1 Initialize fidelity
The initial fidelity test method includes the following steps.
a) Initialize the sub-bit to be measured to the 0 > state.
b) Measure the state of the qubit, sample Nt times, and obtain the number of times it is in the 0 > state as N0.
c) Calculate the probability P0 of initializing each bit to the 0 > state using formula (7).
6.3.7 Energy Consumption
6.3.7.1 No-load energy consumption
The no-load energy consumption test method includes the following steps.
a) The computing system under test runs a standard computing power test program once, such as a random line sampling test program, requiring all quantities to be tested.
The sub-qubit width is measured, and a readout operation is performed every 20 layers of quantum circuit depth to ensure that the quantum computing system is in normal operation.
b) Without changing the configuration of the computing system of the sub-measurement, run it continuously under no-load for 1 hour, measuring and recording every 10 minutes during this period. Primary system energy consumption;
c) After the test is completed, without changing the configuration of the computing system under test, run the standard computing power test program again, for example... The random circuit sampling test procedure requires a readout operation every 20 layers of quantum circuit depth, covering the entire width of the qubits.
This ensures that the quantum computing system operates normally throughout the entire energy consumption test.
d) Take the average value of the 6 recorded power consumption values to obtain the average no-load power consumption.
6.3.7.2 Operating Energy Consumption
The energy consumption test method includes the following steps.
a) The quantum computing system runs standard computing power test procedures, such as random circuit sampling test procedures, requiring the entire qubit width to be tested.
The system performs a readout operation every 20 layers of quantum circuit depth, continuously testing for 1 hour, and measuring system power consumption every 10 minutes.
b) After the test, the average value of the 6 energy consumption values is taken to obtain the average operating energy consumption.
7 Test Report
The performance test report of a quantum computing system should include the following.
a) Test basis;
b) Sample name;
c) The submitting unit and its contact person;
d) Test content and process;
e) Test results and necessary limitations;
f) Sample testing time;
g) Signatures of the tester and reviewer;
h) Name and address of the testing unit;
i) Test report page number.
......
This preview omits tables, figures, formulas and parts of the technical clauses. The complete document — 31 pages — is available in the English PDF.
How to Download GB/T 47468-2026
- 1
Add to cart
Click "Download PDF" on this page and choose "Buy this standard". You can add more standards before checkout.
- 2
Checkout
Enter your email and billing details. Payment is processed securely by Stripe (cards or bank transfer). Prefer bank details? Request a quotation (min. order $300).
- 3
Fast delivery
After payment you'll receive an email confirmation. The English PDF is sent to the same address, usually within minutes and at the latest within 1 business day.
- 4
Invoice included
A tax invoice is attached to the confirmation email. Need a custom invoice? Contact us.
Related Standards
GB/T 10003-2026 — Biaxially oriented polypropylene (BOPP) film
GB/T 10004-2026 — Plastic/Plastic laminated films and pouches for packaging
GB/T 10007-2026 — Rigid cellular plastics - Determination of shear properties
GB/T 47468-2026
$470.00