Talk:List of quantum logic gates

Number of page views in the past 30 days

Improvement notes

Here are some suggestion for improvement.

• Clean-up the text.
• Create more pictures for the diagrams of some of the missing gates.
• Consider adding other gates if the appropriate sources is provided (here I am looking for a book that considers them important not just Qiskit documentation):${\sqrt {Y}},W=(X+Y)/{\sqrt {2}},\mathrm {control} -Y,{\sqrt {H}},\mathrm {control} -H,\cdots$
• Consider adding the 9 qutrit rotation gates, if appropriate source is provided.
• Add more group properties.

--ReyHahn (talk) 18:26, 12 September 2022 (UTC) * Consider moving global phase up to Identity and global phase, as they have the same nature (can be adaptable to any number of qubits and essentially do nothing).--ReyHahn (talk) 07:47, 14 September 2022 (UTC)

The Mølmer–Sørensen gate has different definitions in Cirq, IBM and original papers. I hope that somebody has a good resource that clears this up.--ReyHahn (talk) 07:05, 15 September 2022 (UTC)

@Geek3: I saw your edit, is Mølmer Sørensen not a gate? It is called a gate and has several version check the Mølmer–Sørensen gate talk maybe you can help solve that issue out. Maybe the title can be changed to Mølmer–Sørensen process or something like that.--ReyHahn (talk) 16:35, 25 September 2022 (UTC)

@ReyHahn: Thanks for the remark. The Mølmer Sørensen is not just an abstract logic gate, it is a specific implementation of such a gate on ions. However, I have to correct myself: It is not a CNOT, but as you correctly listed on the other discussions page, either an Ising XX or YY gate (which I guess can be turned into a CNOT by single-qubit operations). I would then follow the original publication (eq. 5) and take it as a YY gate. --Geek3 (talk) 20:13, 25 September 2022 (UTC)

Good idea! For the time being I agree. I hope to find a source that discusses the definition of these MS gates very clearly. I think there is an overlap between what we are calling here "Ising gates" (we still need a good source for those) and the MS gates. Maybe we could add the MS implementation to Rxx as well. What about changing the name of the MS gate article? --ReyHahn (talk) 07:35, 26 September 2022 (UTC)

Let's not change the name of the MS gate article, because MS gate is what it's called by everybody. The word gate has in a way several meanings, of an abstract logic gate and a more specific platform-dependent implementation, as is the case for many other words. I don't see a problem of confusion if we keep the naming. --Geek3 (talk) 20:31, 26 September 2022 (UTC)

Single qubit

@Omnissiahs hierophant: I saw that you added a "Citation needed" to this claim: https://en.wikipedia.org/w/index.php?title=List_of_quantum_logic_gates&type=revision&diff=1110779452&oldid=1110748718. Do you doubt the claim? or do you want to be extra careful and add a reference anyway.--ReyHahn (talk) 21:33, 17 September 2022 (UTC)

It is clear that this is true, for me. But we should add either a source or the tensor product thing that proves it, imho..? · · · Omnissiahs hierophant (talk) 13:21, 19 September 2022 (UTC)

Hermitian

Several gates marked here as hermetian do not seem to be hermitian, for instance the DCNOT gate. They are not self-adjoint. Their transposed matrix is not equal to their complex conjugate. --Geek3 (talk) 11:15, 25 September 2022 (UTC)

You are right. Let me check them again.--ReyHahn (talk) 12:53, 25 September 2022 (UTC)

Phase gate diagram

@Geek3: if you are for the task. In Williams' book there are many diagrams of the phase gate. It is just a simple square with ${\text{Ph}}(\delta )$ written on it, just like most single qubit gates.--ReyHahn (talk) 21:14, 26 September 2022 (UTC)

Well I tried to make it, but the borders and size are weird. Do not hesitate to update it.--ReyHahn (talk) 21:51, 28 September 2022 (UTC)

Ising gates

I am trying to find a good source for Ising gates. Some people call it Ising, some call it Heisenberg and some even Mølmer-Sørensen gates. I think we should change the name to "two qubit rotation gates" or "two qubit interaction gates".--ReyHahn (talk) 08:52, 3 October 2022 (UTC)

periods or intervals

This is maybe not so important, BUT... isn't it better to speak of periods rather than intervals for when the parameterized gates? It is possible to rotate a thing (e.g. an orange, or electron) many many turns. $e^{i\alpha }$ is defined for all $\alpha$, not just [0-2pi). Also I noticed that swap^a and "Core entangling" gates are defined for all real numbers, which deviates from rest of the article. · · · Omnissiahs hierophant (talk) 17:13, 4 October 2022 (UTC)

Quick answer (I am busy this week): agreed it is better to add the period, instead of the interval.--ReyHahn (talk) 09:07, 5 October 2022 (UTC)

Ok, I am basically always overworked/tired nowadays also, but I will change it eventually then.. unless someone else does it or objects :) There is one benefit to the intervals, in that I think that is maybe more concise, and giving periods for all these variables might require more text. Maybe its possible to figure out how to do it beautifully · · · Omnissiahs hierophant (talk) 19:15, 5 October 2022 (UTC)

Determinants

@Omnissiahs hierophant: I am not a fan of the complex determinants, these are unitary matrices these matrices will always have a determinant that is exp(i r) for some real r. The fact of having determinant 1 is nice because it means it is special unitary. Maybe we should just indicate special unitary matrices and not -1 (SU+reflection), i or any of the rest.--ReyHahn (talk) 13:24, 13 October 2022 (UTC)

I contemplated the usefulness of that information, it is probably quite low. I was just filling in the gaps. But, I don't understand why one should leave it out. Is it the usefulness of the information that you mean? Thanks (to speed up communication, i am ok with removing them!) · · · Omnissiahs hierophant (talk) 21:32, 13 October 2022 (UTC)

Tensor gates?

An IP added tensor gates and a number of scalar/vector gates [1] of WP:UNDUE importance. If somebody wants to add this back, please justify it using references. ReyHahn (talk) 11:18, 28 September 2023 (UTC)

Single-Step Parity Check Gate Set for Quantum Error Correction

SSPC gates also can be added here. These gates have been introduced recently and provides to make parity check circuits in one step. It is also shown that these gates are natural for silicon spin qubits and can be implemented as one unitary rather than sequences of cnot. Here is the paper: https://arxiv.org/abs/2306.08849 138.25.4.59 (talk) 23:23, 24 October 2023 (UTC)

This seem more like algorithms than simple gates. Also WP:TOOSOON?--ReyHahn (talk) 23:54, 24 October 2023 (UTC)

no, it is not. Page 28 shows the unitaries of these gates. All you need to do is implement these two unitaries instead of sequential cnot gates. 138.25.4.59 (talk) 23:56, 24 October 2023 (UTC)

+ The identity gate can be thought of as an algorithm, since it is the most difficult gate to implement. To this day, there has been no single experimental demonstration showing us an implementation of identity gates. Doing nothing counts as an identity gate, but it introduces a lot of noise. Implementing an identity gate is a very complicated quantum control algorithm in a sense, but it still is a gate and it exists. 138.25.4.59 (talk) 00:19, 25 October 2023 (UTC)

Adding GPi and GPi2 gates

Do you think it is a good idea to add to the article native gates like GPi or GPi2 like they use in IonQ systems ? ChaseYUL99 (talk) 03:57, 20 February 2024 (UTC)