Primitive Quantum Gates for an SU(3) Discrete Subgroup: sigma(216 times3)

arXiv:2611.08161 · hep-lat, quant-ph · Submitted 2026-11-05 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: High Energy Physics: Lattice. The day in lattice field theory: what the simulations put on the table.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Primitive Quantum Gates for an SU(3) Discrete Subgroup".

Mira: Here's the gist of the paper. We construct a primitive gate set for the digital quantum simulation of the 648-element discrete subgroup of SU(3), Σ(216 × 3) This work provides qubit and qutrit–qubit decompositions for five algorithmic primitives: inversion, group multiplication, trace, group Fourier transform, and kinetic-phase gates The estimated fault-tolerant costs for a fiducial shear-viscosity calculation are 2.3–6.8 × 10 10 T gates in the qubit architecture and 1.3–5.9 × 10 12 R gates in the qutrit–qubit architecture Group Structure and Encoding The paper focuses on the discrete subgroup approximation (dsa) which replaces the continuous gauge group by one of its discrete subgroups, enabling a mapping directly to finite Hilbert spaces while preserving a remnant of gauge symmetry analogous to the lattice discretization of SO(4) The dsa replaces real number arithmetic with modular arithmetic and enables embedding with quantum error correction (qec) schemes For SU(3), there are only four choices for the crystal-like subgroup, including Σ(216 × 3) which has 648 elements The elements of Σ(216 × 3) can be encoded into a register of 11 qubits A mixed encoding scheme is used where each group element g⟩ is encoded as g⟩ = p⟩3q⟩3r⟩3s⟩2t⟩2u⟩2y»3, which suits the group arithmetic Primitive Gate Decompositions The five required primitive gates for lattice gauge theories are the inversion (U−1), multiplication (U×), trace (UTr), Fourier transform (UF), and kinetic-phase gate (UP h) The basic qubit gate set involves standard RZ, Clifford, and controlled-NOT gates, where an arbitrary single qubit gate can be decomposed into at most 3 RZ rotations For fault tolerance each RZ is approximated by Clifford+T gates Resource Estimation and Comparison A resource estimation for both ft qubit and mixed circuits is provided, showing the required T gates and R gates for the fiducial shear-viscosity calculation The mixed encoding shows the qutrit synthesis dominates the accounting, with 1.3 × 10 12 (HKS) to 5.9 × 10 12 (HI) metaplectic R gates coming almost entirely from the T(3) content of U× The mixed route wins only when a native qutrit R is about 10 squared cheaper than a qubit T Implementation Details The inversion gate (U−1) is implemented through a sequence of propagation stages using the ordered product form given in Eq.

Kai: (3), involving multiple applications of U(n−1)−1 gates and controlled multiplications on registers The multiplication gate (U×) extends this by performing a controlled map taking g to hg, which involves updating the registers according to specific rules derived from the inversion stages Fourier Transform and Kinetic-Phase Gate The Fourier transform UF F T is built recursively along the chain of subgroups generated by the sgs ordered product of Eq.

Mira: (3) The kinetic-phase gate UP h then multiplies each irrep state by its kinetic eigenvalue λρ, where λρ is the eigenvalue of HE on the irrep ρ The implementation of UF F T involves a fast Fourier transform (fft) UF F T which reduces the cost from O(G2) to O(G log G) The work constructs the complete set of five primitive gates for Σ(216 × 3), the second-largest crystal- The mixed encoding introduced here removes nearly all of the qubit cost of the group multiplication and inversion, whose Toffoli decompositions are replaced by native qutrit permutations The sgs encoding also permits grouping several registers into larger qudits Exploring such encodings together with compatible error-correcting codes could reveal further resource tradeoffs.

Kai: Appendix Details The appendix provides detailed decompositions for the controlledsum gates, showing how they are implemented on qubit and native qutrit hardware The cost of these gates is tabulated in Table XVII, detailing the resource model based on T(3), E(2,3), and E(3,2) The kinetic-phase gate is drawn in the qubit encoding in Figs.

Title and authors: Kai: So we’re starting with the paper "Primitive Quantum Gates for an SU(three) Discrete Subgroup: sigma(two hundred sixteen times3)" and the authors are Erik Gustafson, Henry Lamm, Edison Murairi, and Sebastian Osorio Perez <ref:2611.08161#pg1,Primitive Quantum Gates for an SU(3) Discrete Subgroup>. They’re working on a discrete subgroup of SU(three), specifically one with six hundred forty-eight elements <ref:2611.08161#pg1,discrete subgroup of SU(3>.

Lev: That ordered product form is what lets them implement the inversion gate, which involves a sequence of propagation stages using the ordered product form given in equation three, involving multiple applications of those U minus one gates and controlled multiplications across registers.

Kai: So they have all five primitives—inversion, multiplication, trace, Fourier transform, and kinetic-phase gates—all ready to go for lattice gauge theories. That’s a complete set of tools for constructing any gauge-invariant term in the Hamiltonian limit.

The paper's summary: Mira: Building on what we just heard about the gate set, the paper's summary explains that they constructed this primitive gate set for SU(three) sigma(two hundred sixteen times3) as a way to achieve digital quantum simulation of lattice gauge theories <ref:2611.08161#pg1>.

Kai: They are basically saying they’ve provided qubit and qutrit–qubit decompositions for all five of these primitives.

Lev: It really boils down to them having the necessary tools—the inversion, group multiplication, trace, Fourier transform, and kinetic-phase gates—for constructing any gauge-invariant term from these primitives in the Hamiltonian limit.

Mira: They also highlight how they use a discrete subgroup approximation to replace continuous gauge groups with finite ones so they can map it directly onto finite Hilbert spaces while retaining some remnant of gauge symmetry similar to lattice discretization of SO(four) <ref:2611.08161#pg2>.

Kai: So, the takeaway is that they’ve provided qubit and qutrit–qubit decompositions for all five of these primitives.

Lev: And the encoding scheme is clever. They take those six hundred forty-eight group elements and encode them into a register of eleven qubits using a mixed encoding scheme where each group element gets represented by a string of powers like one three, two three, etc <ref:2611.08161#pg1>., which fits their specific arithmetic needs.

Mira: It’s important to remember that this works because the dsa allows them to replace real number arithmetic with modular arithmetic and enables embedding with quantum error correction schemes.

Kai: So what they’re doing is providing qubit and qutrit–qubit decompositions for all five of these primitives.

Lev: And they have also shown how to implement a trace gate operation via U G squish, which maps group elements to their trace classes and applies phase rotations based on those classes using a classifier Cls, an adder plus p, and an encoder Enc.

Mira: It’s important that we remember that the dsa allows them to replace real number arithmetic with modular arithmetic and enables embedding with quantum error correction schemes.

Mira: So the takeaway is that they’ve provided qubit and qutrit–qubit decompositions for all five of these primitives.

The paper's improvements: Kai: Now we look at the specific improvements they suggest in their work, moving beyond just stating what they do to explaining *how* it’s better than what was there before.

Mira: They are focusing on how the mixed encoding helps them reduce the cost of group multiplication and inversion by replacing those Toffoli decompositions with native qutrit permutations.

Lev: That’s a big deal because it means they leverage the structure of the group to significantly cut down on the computational load when we implement those gates compared to purely qubit implementations.

Kai: I mean, that’s leveraging structure instead of just brute force decomposition, which is a key theme in quantum simulation research.

Mira: It allows them to use native qutrit permutations instead of those Toffoli decompositions for the group multiplication gate, and this is what makes the mixed route competitive with qubit architectures.

Lev: They’ve also shown how to optimize group multiplication by sharing intermediate computations between primitive gates and measurement-assisted uncomputation, which targets the primary driver of that mixed R gate cost in these simulations.

Kai: That optimization seems like it’s a practical step toward making these simulations run faster on real hardware.

Mira: And they also show how the kinetic-phase gate, UP h is implemented via a fast Fourier transform or fft that reduces complexity from order G squared to order G log G, which is what gives them efficiency there.

Lev: That FFT part is key for scaling up the computation without hitting those quadratic costs too hard.

Kai: So they’re showing how to use the structure of the group to reduce gate counts in a way that's both structural and algorithmic.

Mira: And by showing these different ways to implement things, they are demonstrating that we can find trade-offs between different architectures based on whether a native qutrit R gate is cheap enough compared to a qubit T gate.

Lev: It’s basically saying you can choose the path based on the cost ratio of those two gates.

Kai: So they’re showing how to use the structure of the group to reduce gate counts in a way that's both structural and algorithmic, leading to trade-offs between different architectures based on whether a native qutrit R gate is cheap enough compared to a qubit T gate.

Conclusion: Kai: So we’ve covered the full scope of the paper "Primitive Quantum Gates for an SU(three) Discrete Subgroup: sigma(two hundred sixteen times3)" and how this work provides a complete set of five primitives for simulating lattice gauge theories <ref:2611.08161#pg1,Primitive Quantum Gates for an SU(3) Discrete Subgroup>.

Mira: In short, they’ve given us qubit and qutrit–qubit decompositions for inversion, group multiplication, trace, Fourier transform, and kinetic-phase gates.

Lev: And the overall numbers suggest that the mixed encoding route is competitive if you are willing to invest in those specific native gates.

Kai: It’s a solid piece of work that gives us concrete targets for building the next generation of quantum simulation tools.

Mira: They set up the framework for using these discrete subgroups to map gauge theories onto finite quantum systems, which is a solid step toward simulating more complex physics than we could do with continuous groups alone.

Mira: That's all we have time for this paper, and it sets up a framework for using these discrete subgroups to map gauge theories onto finite quantum systems, which is a solid step toward simulating more complex physics than we could do with continuous groups alone.

Lev: For me, it means if we are targeting a fault-tolerant calculation on this scale, we need to be very careful about which architecture—qubit or qutrit—we choose based on these cost ratios they’ve calculated for the fiducial shear-viscosity calculation.

Erik J. Gustafson, * Henry Lamm, × Edison Murairi, ‡ Sebastian Osorio Perez

KBR, Inc. · Fermi National Accelerator Laboratory · Department of Physics, University of Maryland

hep-lat, quant-ph

Submitted: 2026-11-05

Updated: 2026-11-05

Comments: 30 pages, 56 figures

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 79/100

The gist: We construct a primitive gate set for the digital quantum simulation of the 648-element discrete subgroup of SU(3)…

Key concepts

Discrete Subgroup Approximation (DSA)
This method replaces the continuous SU(3) gauge group with a finite, discrete subgroup, like Σ(216 × 3). This allows complex gauge theory calculations to be mapped directly onto finite quantum Hilbert spaces. It uses modular arithmetic instead of real numbers and is compatible with quantum error correction schemes.
Mixed Encoding Scheme
Group elements are encoded into a mixed register using qubits and qudits, such as $|g angle = |p angle^3|q angle^3|r angle^3|s angle^2|t angle^2|u angle^2$. This specific encoding is chosen because it simplifies the arithmetic required for group operations and allows for efficient implementation of the necessary gates.
Primitive Gate Set
The paper defines five fundamental gates—inversion, multiplication, trace, Fourier transform, and kinetic-phase gate—needed to build any gauge-invariant term in a lattice gauge theory Hamiltonian. These are decomposed into basic qubit rotations (like RZ) and higher-order gates (Clifford+T) for fault tolerance.
Resource Estimation
The study calculates the computational cost of simulating a shear-viscosity calculation using both qubit and mixed architectures. The results show that while the qubit architecture requires $10^{10}$ to $10^{12}$ T gates, the mixed qutrit-qubit architecture can be more efficient in certain operations, depending on the relative cost of native qutrit gates versus qubit T gates.

Terminology

Summary

We construct a primitive gate set for the digital quantum simulation of the 648-element discrete subgroup of SU(3), Σ(216 × 3) This work provides qubit and qutrit–qubit decompositions for five algorithmic primitives: inversion, group multiplication, trace, group Fourier transform, and kinetic-phase gates The estimated fault-tolerant costs for a fiducial shear-viscosity calculation are 2. p. 0

Where the paper says it

at NASA Ames Research Center, Moffett Field, California, 94035, USA 2Fermi National Accelerator Laboratory, Batavia, Illinois, 60510, USA 3Department of Physics, University of Maryland, College Park, MD 20742, USA We construct a primitive gate set for the digital quantum simulation of the 648-element discrete subgroup of SU(3), Σ(216 × 3). The five algorithmic primitives are the inversion, group multiplication, trace, group Fourier transform, and kinetic-phase gates, for which we provide qubit and qutrit–qubit decompositions.

3–6.8 × 10 10 T gates in the qubit architecture and 1. p. 0
Where the paper says it

For a fiducial shear-viscosity calculation on a 103 spatial lattice with 50 time steps and a total synthesis-error budget of 10−8 , the estimated fault-tolerant costs are 2. 8 × 1010 T gates in the qubit architecture and 1.

3–5.9 × 10 12 R gates in the qutrit–qubit architecture

Group Structure and Encoding

The paper focuses on the discrete subgroup approximation (dsa) which replaces the continuous gauge group by one of its discrete subgroups, enabling a mapping directly to finite Hilbert spaces while preserving a remnant of gauge symmetry analogous to the lattice discretization of SO(4) The dsa replaces real number arithmetic with modular arithmetic and enables embedding with quantum error correction (qec) schemes For SU(3), there are only four choices for the crystal-like subgroup, including Σ(216 × 3) which has 648 elements The elements of Σ(216 × 3) can be encoded into a register of 11 qubits A mixed encoding scheme is used where each group element g⟩ is encoded as g⟩ = p⟩3q⟩3r⟩3s⟩2t⟩2u⟩2y»3, which suits the group arithmetic

Primitive Gate Decompositions

The five required primitive gates for lattice gauge theories are the inversion (U−1), multiplication (U×), trace (UTr), Fourier transform (UF), and kinetic-phase gate (UP h) The basic qubit gate set involves standard RZ, Clifford, and controlled-NOT gates, where an arbitrary single qubit gate can be decomposed into at most 3 RZ rotations For fault tolerance each RZ is approximated by Clifford+T gates

Resource Estimation and Comparison

A resource estimation for both ft qubit and mixed circuits is provided, showing the required T gates and R gates for the fiducial shear-viscosity calculation The mixed encoding shows the qutrit synthesis dominates the accounting, with 1.3 × 10 12 (HKS) to 5.9 × 10 12 (HI) metaplectic R gates coming almost entirely from the T(3) content of U× The mixed route wins only when a native qutrit R is about 10 squared cheaper than a qubit T

Implementation Details

The inversion gate (U−1) is implemented through a sequence of propagation stages using the ordered product form given in Eq. (3), involving multiple applications of U(n−1)−1 gates and controlled multiplications on registers The multiplication gate (U×) extends this by performing a controlled map taking g to hg, which involves updating the registers according to specific rules derived from the inversion stages

Fourier Transform and Kinetic-Phase Gate

The Fourier transform UF F T is built recursively along the chain of subgroups generated by the sgs ordered product of Eq. (3) The kinetic-phase gate UP h then multiplies each irrep state by its kinetic eigenvalue λρ, where λρ is the eigenvalue of HE on the irrep ρ The implementation of UF F T involves a fast Fourier transform (fft) UF F T which reduces the cost from O(G2) to O(G log G)

Conclusion and Outlook

The work constructs the complete set of five primitive gates for Σ(216 × 3), the second-largest crystal- The mixed encoding introduced here removes nearly all of the qubit cost of the group multiplication and inversion, whose Toffoli decompositions are replaced by native qutrit permutations The sgs encoding also permits grouping several registers into larger qudits Exploring such encodings together with compatible error-correcting codes could reveal further resource tradeoffs.

Appendix Details

The appendix provides detailed decompositions for the controlledsum gates, showing how they are implemented on qubit and native qutrit hardware The cost of these gates is tabulated in Table XVII, detailing the resource model based on T(3), E(2,3), and E(3,2) The kinetic-phase gate is drawn in the qubit encoding in Figs. 16 and 17 of Sec. p. 1

Where the paper says it

Important mathematical Σ(216 × 3) is the qutrit Clifford group, a link between the dsa and qec. properties of Σ(216 × 3) are summarized in Sec.

VIII.1 The cost of UΣ(23×3) F F T is detailed in Table IX, showing the total resource counts for both architectures

Overall Cost Analysis

The per-link-per-τ gate counts follow from the pergate costs (Tabs. XI, XII) and the per-link multiplicities (Tab. p. 0

Where the paper says it

Other strategies reformulate the theory in a different set of variables or basis, such as a q-deformed formulation that replaces the gauge group with a corresponding quantum group [59– 61], the loop-string-hadron framework [62–67], light-front quantization [68–70], a magnetic or mixed basis [71–73], maximal-tree gauge fixing [74], conformal truncation [75], strong-coupling and large-Nc expansions [76, 77], fuzzy gauge theories [78], an orbifold digitization [79–85], and quantum link models [86–96]. Across this diversity of approaches, most find that ∼ 10 qubits per SU(3) degree of Super Intelligence Usage Disclosure This document was created with assistance from Anthropic’s Claude and OpenAI’s Chat GPT for generating text and code.

XIII) The total ft resource counts for this fiducial calculation are summarized in Table XV The improved-Hamiltonian qubit cost is nearly six times the FFT-based estimate for Σ(36 × 3) The Fourier transform carries about half of the qubit cost, the trace and multiplication gates a fifth each The mixed encoding splits the cost into the qubit-T side channel, the qutrit R channel, and the two mixed entanglers The total ft resources for this fiducial calculation are 2. p. 0
Where the paper says it

The five algorithmic primitives are the inversion, group multiplication, trace, group Fourier transform, and kinetic-phase gates, for which we provide qubit and qutrit–qubit decompositions. For a fiducial shear-viscosity calculation on a 103 spatial lattice with 50 time steps and a total synthesis-error budget of 10−8 , the estimated fault-tolerant costs are 2.

3 × 10 10 (HKS) to 6.8 × 10 10 (HI) T gates The mixed entanglers E(2,3) and E(3,2) contribute 5. p. 3
Where the paper says it

For two qubits this is the usual cnot, CaXb = C 1 aX (0,1) b , and we suppress the labels there. When counting mixed resources, the two mixed cases are tallied separately as the entanglers E(d,f) , whose subscripts are the control and target dimensions d, f ∈ {2, 3}: E(2,3) (qubit control, qutrit target) and E(3,2) (qutrit control, qubit target) (see App.

1 × 10 8 to 2.1 × 10 9 and 7.4 × 10 7 to 1.8 × 10 8 The mixed R-count exceeds the qubit T-count by a factor of 55 to 90, so the mixed route wins only when a native qutrit R is about two orders of magnitude cheaper than a qubit T. p. 2
Where the paper says it

C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13 C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 C24 ω aC1 CC ω aCV 2 ω aCV CED CED2 ω aCD ω aCD2 ω aCV 2D ω aCV D2 Size 1 1 1 24 9 9 9 54 54 54 72 72 12 12 12 12 12 12 36 36 36 36 36 36 Ord. 1 3 3 3 2 6 6 4 12 12 3 3 9 9 9 9 9 9 18 18 18 18 18 18 1 (0) 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 (1) 1 1 1 1 1 1 1 1 1 1 ω ω 2 ω ω ω ω 2 ω 2 ω 2 ω ω ω ω 2 ω 2 ω 2 1 (2) 1 1 1 1 1 1 1 1 1 1 ω 2 ω ω 2 ω 2 ω 2 ω ω ω ω 2 ω 2 ω 2 ω ω ω 2 (0) 2 2 2 2 -2 -2 -2 0 0 0 -1 -1 -1 -1 -1 -1 -1 -1 1 1 1 1 1 1 2 (1) 2 2 2 2 -2 -2 -2 0 0 0 -ω -ω 2 -ω -ω -ω -ω 2 -ω 2 -ω 2 ω ω ω ω 2 ω 2 ω 2 2 (2) 2 2 2 2 -2 -2 -2 0 0 0 -ω 2 -ω -ω 2 -ω 2 -ω 2 -ω -ω -ω ω 2 ω 2 ω 2 ω ω ω 3 (a) 3 3 3 3 3 3 3 -1 -1 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 (0) 3 3ω 3ω 2 0 -1 -ω -ω 2 1 ω ω2 0 0 ω 2µ ∗ µ ∗ ωµ∗ µ ωµ ω2µ -ζ 2 -ζ 4∗ -ζ ∗ -ζ 2∗ -ζ -ζ 4 3 (1) 3 3ω 3ω 2 0 -1 -ω -ω 2 1 ω ω2 0 0 µ ∗ ωµ∗ ω 2µ ∗ ω 2µ µ ωµ -ζ 4∗ -ζ ∗ -ζ 2 -ζ 4 -ζ 2∗ -ζ 3 (2) 3 3ω 3ω 2 0 -1 -ω -ω 2 1 ω ω2 0 0 ωµ∗ ω 2µ ∗ µ ∗ ωµ ω2µ µ -ζ ∗ -ζ 2 -ζ 4∗ -ζ -ζ 4 -ζ 2∗ 3 (0)∗ 3 3ω 2 3ω 0 -1 -ω 2 -ω 1 ω 2 ω 0 0 ωµ µ ω2µ µ ∗ ω 2µ ∗ ωµ∗ -ζ 2∗ -ζ 4 -ζ -ζ 2 -ζ ∗ -ζ 4∗ 3 (1)∗ 3 3ω 2 3ω 0 -1 -ω 2 -ω 1 ω 2 ω 0 0 µ ω2µ ωµ ωµ∗ µ ∗ ω 2µ ∗ -ζ 4 -ζ -ζ 2∗ -ζ 4∗ -ζ 2 -ζ ∗ 3 (2)∗ 3 3ω 2 3ω 0 -1 -ω 2 -ω 1 ω 2 ω 0 0 ω 2µ ωµ µ ω 2µ ∗ ωµ∗ µ ∗ -ζ -ζ 2∗ -ζ 4 -ζ ∗ -ζ 4∗ -ζ 2 6 (0) 6 6ω 6ω 2 0 2 2ω 2ω 2 0 0 0 0 0 -ωµ∗ -ω 2µ ∗ -µ ∗ -ωµ -ω 2µ -µ -ζ ∗ -ζ 2 -ζ 4∗ -ζ -ζ 4 -ζ 2∗ 6 (1) 6 6ω 6ω 2 0 2 2ω 2ω 2 0 0 0 0 0 -ω 2µ ∗ -µ ∗ -ωµ∗ -µ -ωµ -ω 2µ -ζ 2 -ζ 4∗ -ζ ∗ -ζ 2∗ -ζ -ζ 4 6 (2) 6 6ω 6ω 2 0 2 2ω 2ω 2 0 0 0 0 0 -µ ∗ -ωµ∗ -ω 2µ ∗ -ω 2µ -µ -ωµ -ζ 4∗ -ζ ∗ -ζ 2 -ζ 4 -ζ 2∗ -ζ 6 (0)∗ 6 6ω 2 6ω 0 2 2ω 2 2ω 0 0 0 0 0 -ω 2µ -ωµ -µ -ω 2µ ∗ -ωµ∗ -µ ∗ -ζ -ζ 2∗ -ζ 4 -ζ ∗ -ζ 4∗ -ζ 2 6 (1)∗ 6 6ω 2 6ω 0 2 2ω 2 2ω 0 0 0 0 0 -ωµ -µ -ω 2µ -µ ∗ -ω 2µ ∗ -ωµ∗ -ζ 2∗ -ζ 4 -ζ -ζ 2 -ζ ∗ -ζ 4∗ 6 (2)∗ 6 6ω 2 6ω 0 2 2ω 2 2ω 0 0 0 0 0 -µ -ω 2µ -ωµ -ωµ∗ -µ ∗ -ω 2µ ∗ -ζ 4 -ζ -ζ 2∗ -ζ 4∗ -ζ 2 -ζ ∗ 8 (0) 8 8 8 -1 0 0 0 0 0 0 -1 -1 2 2 2 2 2 2 0 0 0 0 0 0 8 (1) 8 8 8 -1 0 0 0 0 0 0 -ω -ω 2 2ω 2ω 2ω 2ω 2 2ω 2 2ω 2 0 0 0 0 0 0 8 (2) 8 8 8 -1 0 0 0 0 0 0 -ω 2 -ω 2ω 2 2ω 2 2ω 2 2ω 2ω 2ω 0 0 0 0 0 0 9 9 9ω 9ω 2 0 -3 -3ω -3ω 2 -1 -ω -ω 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9 ∗ 9 9ω 2 9ω 0 -3 -3ω 2 -3ω -1 -ω 2 -ω 0 0 0 0 0 0 0 0 0 0 0 0 0 0 without native qutrits we also use a qubit-only encoding, mapping each qutrit to two qubits: |x⟩3 → |x1⟩2|x2⟩2 (6) where x = 2x1 + x2.

Future Work

With Σ(216 × 3) complete, only Σ(360 × 3), the largest crystallographic subgroup of SU(3), remains before the discrete-subgroup tower of SU(3) is exhausted Beyond the pure-gauge sector, faithful QCD needs fermion fields discretized as staggered or Wilson fermions The systematic and renormalization uncertainties intrinsic to the discrete-group approximation remain the central open questions.

Acknowledgments

EG is a KBR employee working under the Intelligent Systems Research and Development-3 (ISRDS-3) Contract 80ARC020D0010 EG, HL, and EM were supported by the U.S. Department of Energy, Office of Science, National Quantum Information Science Research Centers, Superconducting Quantum Materials and Systems Center (SQMS) under Contract No. p. 0

Where the paper says it

Gustafson,1, ∗ Henry Lamm,2, † Edison Murairi,2, ‡ and Sebastian Osorio Perez2, 3, § 1KBR, Inc. at NASA Ames Research Center, Moffett Field, California, 94035, USA 2Fermi National Accelerator Laboratory, Batavia, Illinois, 60510, USA 3Department of Physics, University of Maryland, College Park, MD 20742, USA We construct a primitive gate set for the digital quantum simulation of the 648-element discrete subgroup of SU(3), Σ(216 × 3).

89243024CSC000002 Claude and ChatGPT were used to assist with circuit implementation, drafting, and editorial reviews.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved simulation of lattice gauge theories (lgts) by constructing primitive quantum gates for SU(3) discrete subgroups, specifically Σ(216 × 3), enabling digital quantum simulation of these theories. The system can execute fiducial shear-viscosity calculations on a 103 spatial lattice with 50 time steps, achieving estimated fault-tolerant costs of 2.3–6.8 x 10 10 T gates in the qubit architecture or 1.3–5.9 x 10 12 R gates in the qutrit-qubit architecture.

  2. Implementation of a complete five-gate primitive gate set for lattice gauge theories, which includes inversion, group multiplication, trace, group Fourier transform, and kinetic-phase gates with qubit and qutrit-qubit decompositions. This allows for the construction of any gauge-invariant term from these primitives in the Hamiltonian limit.

  3. Efficient encoding of group elements using a strong generating set (sgs) to map the group Σ(216 × 3) onto a mixed register of qubits and qudits, which suits the group arithmetic for gate operations. This mixed encoding allows for an ordered product form that corresponds to a series of finite groups, enabling direct mapping to Hilbert spaces while preserving a remnant of gauge symmetry.

  4. Development of fault-tolerant gate decompositions for qubit and qutrit architectures, showing costs in terms of T gates and R gates, which are crucial for resource estimation in error-corrected computations. The system can utilize Clifford+T gates to approximate arbitrary single qubit rotations and decompose higher-order controlled operations into exact nested staircases at the cost of clean ancillas.

  5. Construction of a mixed-register circuit for the group multiplication gate U Σ(216 × 3) ×, which utilizes controlled-sum primitives and an ancilla flag to manage register updates, reducing the cost compared to purely qubit implementations by leveraging native qutrit permutations. This circuit reuses stage gates from the inversion algorithm.

  6. Implementation of a trace gate operation via U G squish, which maps group elements to their trace classes and applies phase rotations based on these classes. This operation uses a classifier Cls, an adder +p, and an encoder Enc to compute the required information for phase application in both qubit and mixed architectures.

  7. Construction of the Fourier transform gate UF F T recursively along the chain of subgroups generated by the sgs ordered product, enabling an efficient implementation via nested series of subgroups. This allows for a fast Fourier transform (fft)UF F T that reduces complexity from O(G 2) to O(G log G) and incorporates kinetic-phase gates.

  8. Synthesis of arbitrary single-qutrit SU(3) gates by decomposing them into nine two-level rotations R(a,b) alpha with alpha in the set X, Y, Z. This decomposition allows for the synthesis of an arbitrary single-qutrit operation from Clifford+R gates and a non-Clifford supplement like the qutrit T gate T(3).

  9. Resource estimation for fault-tolerant circuits by calculating per-primitive share costs for both qubit and mixed encodings, comparing the total cost of the fiducial shear-viscosity calculation on Σ(216 × 3) to FFT-based estimates for smaller groups like Σ(36 × 3). This comparison informs the hardware choice based on the cost ratio of a qutrit R gate to a qubit T gate.

  10. Optimization of group multiplication by sharing intermediate computations between primitive gates and measurement-assisted uncomputation, aiming to reduce its dominant cost in the mixed architecture. This optimization targets the primary driver of mixed R-gate cost in these simulations.

Abstract

We construct a primitive gate set for the digital quantum simulation of the 648-element discrete subgroup of SU(3),. The five algorithmic primitives are the inversion, group multiplication, trace, group Fourier transform, and kinetic-phase gates, for which we provide qubit and qutrit--qubit decompositions. For a fiducial shear-viscosity calculation on a 10 cubed spatial lattice with 50 time steps and a total synthesis-error budget of 10-8, the estimated fault-tolerant costs are 2.3 -- 6.8 times10 10 T gates in the qubit architecture and 1.3 -- 5.9 times10 12 R gates in the qutrit--qubit architecture.

Sources

Related papers