Qubit Quality
- QCR by GQI
- Dec 26, 2017
- 5 min read
Qubit Quality Table 1
Qubit ConnectivityT1 (µsec)T2 (µsec)ComputerQubit CountMinMaxAveMinMaxAveMinMaxAve
IBM Sherbrooke 127 (QV:32, CLOPS:904)
127
1
3
1.126
27.88
547.72
312.408
16.72
474.07
177.53
IBM Washington 127 (QV:64, CLOPS:850)
127
1
3
2.2
16.24
172.35
94.95
1.72
219.11
97.25
IBM Q Brooklyn 65 (QV:32. CLOPS:1.5K)
65
1
3
2.2
17.12
134.68
80.17
12.48
140.63
78.64
IBM Q Kolkata 27 (QV:128, CLOPS:2K)
27
1
3
2.1
9.86
239.52
126.14
11.52
167.87
78.64
IBM Peekskill 27 (QV:TBD, CLOPS: TBD)
27
1
3
2.1
39.43
453.13
266.14
13.98
663.59
256.61
IBM Perth 7 (QV:32, CLOPS:2.9K)
7
1
3
1.7
123.87
279.08
174.27
74.65
211.32
145.20
Rigetti Aspen-11 (CLOPS: >4000) See Note 17
40
2
3
2.4
4.40
116.9
31.13
0.97
29.74
11.88
Rigetti Aspen-M-2 (CLOPS: >4000) See Note 17
80
2
3
2.4
3.50
166
34.81
0.51
94.27
22.77
Rigetti Aspen-M-3 (CLOPS: N/A) See Note 17
80
2
3
2.4
3.7
58.7
22.1 Median 25.0 Mean
1.5
76.2
24.2 Median 28.0 Mean
AQT 24 Qubit
24
24
24
24
N/A
N/A
>106
N/A
N/A
>105
IonQ Harmony
11
11
11
11
>1010
>1010
>1010
TBD
TBD
~3x106
IonQ Aria
<=32
All
All
All
10 sec.
100 sec.
TBD
N/A
N/A
~ 1 sec.
Quantinuum (Honeywell) H1-1 20 Qubit18 (QV32768)
20
20
20
20
TBD
TBD
>106 sec.
TBD
TBD
~3 sec.
Quantinuum (Honeywell) H1-2 12 Qubit18 (QV4096)
12
12
12
12
TBD
TBD
>106 sec.
TBD
TBD
~3 sec.
USTC Zuchongzhi 2.0
66
1
4
3.33
15.6
46.6
30.6
2.0
16.0
5.3
USTC Zuchongzhi 2.1
66
1
4
3.33
15
37
26.5
1.1
12.1
3.4
Zhejiang University Processor I
68
1
4
~2.5
N/A
N/A
109.8
N/A
N/A
N/A
Baidu QIAN SHI
10
N/A
N/A
N/A
N/A
N/A
31.0
N/A
N/A
8.7
Google Sycamore
53
1
4
3.25
11
21
15
N/A
N/A
N/A
Google Improved Sycamore for ECC Test
21
1
2
1.9
N/A
N/A
15
N/A
N/A
19
ColdQuanta Hilbert
~100
~4
8
~6
N/A
N/A
N/A
N/A
N/A
N/A
Pasqal 200 Qubit
200
1
N/A
2016
N/A
N/A
Tens of Years
N/A
N/A
Tens of Milliseconds
QuTech Spin-2
2
1
1
1
N/A
N/A
>20,000
N/A
N/A
>6
QuTech Starmon-5
5
1
4
2.4
8.9
24.0
17.5
14.6
26.8
23.0
Qubit Quality Table 2
1-Qubit Gate Fidelity2-Qubit Gate FidelityReadout FidelityComputerMinMaxAveMinMaxAveMinMaxAve
IBM Sherbrooke 127 (QV:32, CLOPS:904)
98.47%
99.99%
99.95%
96.62%
99.68%
99.16%
88.86% 83.84%
99.92% 99.74%
98.46% (Prepare 0, Read 1) 98.16% (Prepare 1, Read 0)
IBM Washington 127 (QV:64, CLOPS:850)
94.17%
99.98%
99.88%
89.03%
99.44%
98.04%
77.20% 27.80%
100% 99.60%
97.34% (Prepare 0, Read 1) 96.05% (Prepare 1, Read 0)
IBM Q Brooklyn 65 (QV:32. CLOPS:1.5K)
99.61%
99.97%
99.94%
89.83%
99.37%
98.91%
92.88% 72.98%
99.56% 98.32%
98.37% (Prepare 0, Read 1) 95.58% (Prepare 1, Read 0)
IBM Q Kolkata 27 (QV:128, CLOPS:2K)
99.73%
99.99%
99.97%
89.31%
99.57%
98.78%
96.06% 93.10%
99.62% 99.34
98.86% (Prepare 0, Read 1) 98.34% (Prepare 1, Read 0)
IBM Peekskill 27 (QV:TBD, CLOPS: TBD)
99.74%
99.99%
99.94%
N/A
N/A
98.98%
88.40%
99.60%
98.33%
IBM Perth 7 (QV:32, CLOPS:2.9K)
99.88%
99.98%
99.95%
96.46%
99.29%
98.14%
98.39%
99.43%
98.79%
Rigetti Aspen-11 (CLOPS: >4000) See Note 17
88.87%
99.89%
98.50% Simultaneous
66.64% 76.94%
98.67% 98.98%
89.63%(XY) 91.01%(CZ)
69.85%
99.00%
89.60% (Readout) 99.1% (Active Reset)
Rigetti Aspen-M-2 (CLOPS: >4000) See Note 17
74.27%
100.00%
98.89% Simultaneous
72.14% 60.55%
99.34% 99.05%
89.71%(XY) 89.44%(CZ)
54.00%
98.80%
92.10% (Readout) 99.8% (Active Reset)
Rigetti Aspen-M-3 (CLOPS: N/A) See Note 17
89.9%
99.9%
99.7% Median 99.3% Mean Simultaneous
68.1% 76.3%
98.9% 99.6%
95.1% (XY) Median 91.0% (XY) Mean 94.7% (CZ) Median 91.2% (CZ) Mean
N/A
N/A
96.2% (Readout) 99.7% (Active Reset)
AQT 24 Qubit
99.38%
99.83%
N/A
N/A
N/A
99.0%
Note 12
Note 12
Note 12
IonQ Harmony
99.18%
99.64%
99.46%
95.1%
98.9%
97.5%
N/A
N/A
99.3%
IonQ Aria
N/A
N/A
99.95%
N/A
N/A
99.6%
N/A
N/A
N/A
Quantinuum (Honeywell) H1-120 Qubit (QV32768)
99.97%
99.999%
99.9955%
99.5%
99.83%
99.795%
99.5%
99.8%
99.69%
Quantinuum (Honeywell) H1-212 Qubit (QV4096)
TBD
TBD
99.994%
TBD
TBD
99.81%
TBD
TBD
99.72%
USTC Zuchongzhi 2.0 Simultaneous15
99.73%
99.92%
99.86%
98.75%
99.71%
99.41%
90.08%
98.66%
95.48%
USTC Zuchongzhi 2.1 Simultaneous15
N/A
N/A
99.84%
N/A
N/A
99.40%
N/A
N/A
97.74%
USTC Zuchongzhi 2.1 Simultaneous15
N/A
N/A
99.84%
N/A
N/A
99.40%
N/A
N/A
97.74%
Zhejiang University Processor I Simultaneous
N/A
N/A
99.91%
N/A
N/A
99.40%
N/A
N/A
N/A
Baidu QIAN SHI
N/A
N/A
99.8%
N/A
N/A
96.40% (CX) 96.80% (CZ)
N/A
N/A
N/A
Google Sycamore Isolated
99.8%
99.9%
99.9%
98.1%
99.3%
99.1%
97.4% 91%14
99.5% 97%14
98.9% (Prepare 0, Read 1) 95% (Prepare 1, Read 0)
Google Sycamore Simultaneous
99.66%
99.92%
99.84%
96.7%
99.2%
98.6%
97% 91%14
99% 97%14
98% (Prepare 0, Read 1) 93% (Prepare 1, Read 0)
Google Improved Sycamore for ECC Test Simultaneous
N/A
N/A
99.89%
N/A
N/A
99.34
N/A
N/A
98.1%
ColdQuanta Hilbert
N/A
N/A
99.1%
N/A
N/A
95%
N/A
N/A
N/A
Pasqal 200 Qubit
N/A
N/A
>99%
98%
99.1%
98.5%
N/A
N/A
99% (Prepare 0, Read 1) 97% (Prepare 1, Read 0)
QuTech Spin-2
N/A
N/A
~99.0%
N/A
N/A
>90%
N/A
N/A
~85%10
QuTech Starmon-5
99.8%
99.9%
99.84%
95.3%
97.9%
97.15%
95.3%
98.7%
97.08%
Qubit Quality Table 3
Qubit Gate Delays
Computer
Min
Max
Average
IBM Perth 7 (QV:32, CLOPS:2.9K)
277 ns.
640 ns.
443 ns. (2-Qubit)
IBM Q Kolkata 27 (QV:128, CLOPS:2K)
196 ns.
1209 ns.
418 ns. (2-Qubit)
IBM Q Washington 127 (QV:64, CLOPS:850)
270 ns.
1344 ns.
547 ns. (2-Qubit)
Rigetti Aspen-11 (CLOPS: >4000)
N/A
N/A
240 ns. (2-Qubit) 40 ns. (Single)
Rigetti Aspen-M-2 (CLOPS: >4000)
N/A
N/A
240 ns. (2-Qubit) 40 ns. (Single)
Rigetti Aspen-M-3 (CLOPS: N/A)
N/A
N/A
240 ns. (2-Qubit) 40 ns. (Single)
USTC Zuchongzhi 2.0
N/A
N/A
32 ns. (2-Qubit)
USTC Zuchongzhi 2.1
N/A
N/A
24 ns. (2-Qubit)
Google Sycamore
N/A
N/A
20-25 ns. (Single) 12-32 ns. (2-Qubit)
Google Improved Sycamore for ECC Test
N/A
N/A
14 ns. (Single) 26 ns. (2-Qubit CZ) 600 ns. (Readout) 280 ns. (Reset)
IonQ Aria
N/A
N/A
135,000 ns. (Single) 600,000 ns. (2-Qubit)
ColdQuanta Hilbert
N/A
N/A
500 ns. (2-Qubit)
AQT 24 Qubit
N/A
N/A
200,000 ns. (2-Qubit)
QuTech Spin-2
N/A
N/A
150 ns. (2-Qubit) 250 ns. (Single)
QuTech Starmon-5
N/A
N/A
60 ns. (2-Qubit) 20 ns. (Single)
Qubit Quality Table 4
ComputerReferenceDate
IBM Sherbrooke 127 (QV:32, CLOPS:904)
https://cloud.ibm.com/quantum/resourceshttps://research.ibm.com/blog/eagle-quantum-error-mitigation
12/24/2022
IBM Washington 127 (QV:64, CLOPS:850)
https://cloud.ibm.com/quantum/resources
2/16/2022
IBM Q Brooklyn 65 (QV:32. CLOPS:1.5K)
https://cloud.ibm.com/quantum/resources
2/16/2022
IBM Q Kolkata 27 (QV:128, CLOPS:2K)
https://cloud.ibm.com/quantum/resources
2/16/2022
IBM Peekskill 27 (QV:TBD, CLOPS: TBD)
https://cloud.ibm.com/quantum/resources
2/16/2022
IBM Perth 7 (QV:32, CLOPS:2.9K)
https://cloud.ibm.com/quantum/resources
2/16/2022
Rigetti Aspen-11 (CLOPS: >4000)
https://qcs.rigetti.com/qpus
8/12/2022
Rigetti Aspen-M-@ (CLOPS: >4000)
https://qcs.rigetti.com/qpus
8/12/2022
Rigetti Aspen-M-3 (CLOPS: >4000)
https://qcs.rigetti.com/qpus/
1/18/2023
AQT 24 Qubit
https://arxiv.org/abs/2101.11390
1/27/2021
IonQ Harmony
https://ionq.co/news/december-11-2018#appendixhttps://arxiv.org/abs/1903.08181
3/19/2018
IonQ Aria
https://ionq.com/posts/july-25-2022-ionq-aria-part-one-practical-performance
8/18/2022
Quantinuum (Honeywell) H1-1 20 Qubit (QV32768)
https://assets-global.website-files.com/617730fbcf7b7c387194556a/62a8f7244596df4d854c2222_Quantinuum%20H1%20Product%20Data%20Sheet%20v5%2014JUN22.pdfhttps://www.quantinuum.com/pressrelease/quantinuum-sets-new-record-with-highest-ever-quantum-volumehttps://www.quantinuum.com/news/quantum-volume-reaches-5-digits-for-the-first-time-5-perspectives-on-what-it-means-for-quantum-computing
6/14/2022 9/27/2022 2/24/2023
Quantinuum (Honeywell) H1-2 12 Qubit (QV4096)
https://www.quantinuum.com/pressrelease/quantinuum-announces-quantum-volume-4096-achievement
https://arxiv.org/abs/2106.14734
6/29/2021
USTC Zuchongzhi 2.1
9/9/2021
Zhejiang University Processor I
Baidu QIAN SHI
https://quantum.baidu.com/
8/25/2022
https://www.nature.com/articles/s41586-019-1666-5https://static-content.springer.com/esm/art%3A10.1038%2Fs41586-019-1666-5/MediaObjects/41586_2019_1666_MOESM1_ESM.pdf
https://quantumai.google/hardware/datasheet/weber.pdf
10/23/2019
Google Improved Sycamore for ECC Test
https://www.nature.com/articles/s41586-021-03588-y
https://www.youtube.com/watch/?v=mmyq1ubjqO8
ColdQuanta Hilbert
https://cdn.sanity.io/files/sbcw1clc/production/2b54cd985fcba6974aa6350f3209447c77121aab.pdf
7/7/2021
3/25/2021
https://qutech.nl/wp-content/uploads/2020/04/3.-Technical-Fact-Sheet-Quantum-Inspire-Starmon-5.pdfhttps://qutech.nl/wp-content/uploads/2020/04/2.-Technical-Fact-Sheet-Quantum-Inspire-Spin-2.pdf
For most of the parameters we show the Min, Max, and Average values. Since IBM publishes the individual values for every qubit, the Min shows the value for the worst of the qubits, the Max shows the value for the best of the qubits, and the Average shows the mean calculations for all of the qubits.
The connectivity shows the number of connections from a qubit to a other qubits in the array for use in creating a CNOT gate. The higher the connectivity, the easier it would be to fit a quantum calculation into the structure. At this time, we do not differentiate on the flexibility of a connection. For example, if qubit 1 is connected to qubit 2, many implementations require one of the qubits to be the CONTROL and the other qubit to be the TARGET. Some implementations may be flexible enough so that either qubit can serve as the CONTROL and either qubit can serve as the TARGET. That implementation may have some configuration advantages, but for the purposes of the table we are still only counting it as one connection and not as two.
The T1 measure is called the relaxation time and the T2 or T2* measure is called the decoherence time. For details of these definitions we refer you to this paper. Note that IBM only publishes the T2 times while Rigetti only publishes the T2* time. The measures are similar, but not exactly the same.
The IBM reference link in Table 4 may require you to register for the IBM Q Experience. If you click on this link it may ask you for a logon and password to see in more detail the referenced data.
The IonQ Readout Fidelity measure of 99.3% includes both state preparation and measurement errors.
Google publishes both isolated and simultaneous gate fidelity numbers. For this table we are showing both the simultaneous numbers and the isolated numbers. The simultaneous numbers are slightly worse, but more realistic, in our opinion. But to provide the best comparison we can, we also show the isolated numbers because we think that is what some of the other platforms are using.
For the Rigetti Aspen-7, the Readout Value is the SPAM (State Preparation and Measurement) value which will always be slightly lower than the pure readout value because it includes any state preparation errors.
This note is no longer used.
For the QuTech Spin-2, the readout numbers include both initialization plus readout.
The IonQ 32 qubit data was taken from a paper published by authors from the University of Maryland, Duke University, and Georgia Institute of Technology. IonQ has indicated that the system mentioned in the paper is nearly identical to their own.
The AQT fidelity numbers includes SPAM (State Preparation and Measurement) errors.
The QV numbers listed for the IBM machines, except IBM Armonk, indicate IBM's Quantum Volume measurement for that machine.
For Google's readout numbers, the first number represents prepare 0 and read 1 fidelity, the second number represents prepare 1 and read 0 fidelity.
The USTC Zuchongzhi 2.0 fidelity numbers are based upon the best 56 qubits from the 66 qubit chip. Zuchongzhi 2.1 fidelity numbers are based upon the best 60 qubits.
The Pasqal devices has a reconfigurable geometry so the Connectivity for a qubit can be configured anywhere between 1 to 20.
Rigetti sometimes quotes these numbers as Median Averages. These tables use Arithmetic Averages.
The Quantinuum H1-1 can do 5 two-qubit operations simultaneously, the H1-2 can do 3 two-qubit operations simultaneously.
Questions, suggestions, and any additions you may have to the data are welcomed. You can send them to info@quantumcomputingreport.com.
Comments