LM implementation of multi-wire gates #358

multiphaseCFD · 2022-09-21T20:41:39Z

Before submitting

Please complete the following checklist when submitting a PR:

All new features must include a unit test.
If you've fixed a bug or added code that should be tested, add a test to the
tests directory!
All new functions and code must be clearly commented and documented.
If you do make documentation changes, make sure that the docs build and
render correctly by running make docs.
Ensure that the test suite passes, by running make test.
Add a new entry to the .github/CHANGELOG.md file, summarizing the
change, and including a link back to the PR.
Ensure that code is properly formatted by running make format.

When all the above are checked, delete everything above the dashed
line and fill in the pull request template.

Context:
This PR add LM implementation for 3 Qubits gates (CSWAP & Toffoli) and QChem 4 Qubits gates (DoubleExcitation, DoubleExcitationMinus, DoubleExcitationPlus).
Description of the Change:
Add LM implementation for 3 Qubits gates (CSWAP & Toffoli) and QChem 4 Qubits gates (DoubleExcitation, DoubleExcitationMinus, DoubleExcitationPlus).
Benefits:
The speed of LM implementation of CSWAP, Toffoli and DoubleExcitation gates is faster than their PI implementation for all tests number of qubits. Per DoubleExcitationMinus and DoubleExcitationPlus gates, LM implementation is faster than PI implementation for small number qubit systems (<10 qubits), while slower than PI implementation for larger number of qubit systems (>10 qubits).

Possible Drawbacks:

Related GitHub Issues:
#315

* Adding new operations file; including in test file; updating Makefile to include pybind path * Adding Paulis and tests * Adding using directives; removing pybind dependence from Makefile and operations.hpp * Renaming test cases * Apply suggestions, change naming conventions * Adding Hadamard * Adding S, T and rotation gates * Update pennylane_lightning/src/operations.hpp Co-authored-by: Tom Bromley <49409390+trbromley@users.noreply.github.com> * Update pennylane_lightning/src/operations.hpp Co-authored-by: Tom Bromley <49409390+trbromley@users.noreply.github.com> * Adding CNOT * Changing local var name for T * Update pennylane_lightning/src/tests/lightning_unittest.cpp Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com> * Update pennylane_lightning/src/tests/lightning_unittest.cpp Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com> * Update pennylane_lightning/src/tests/lightning_unittest.cpp Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com> * Update pennylane_lightning/src/tests/lightning_unittest.cpp Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com> * Transition to using expect_true * Applying Nathan's suggestions on naming from code review * Creating aliases for 1-3 qubit states; using these in the tests * Adding test for three qubit CNOT where the control is the third qubit; correcting 3q rank * updating CNOT test Co-authored-by: Tom Bromley <49409390+trbromley@users.noreply.github.com> Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com>

Adds skeleton for C++ section of lightning.qubit

* factor out permutations * Remove hadamard and add in operations header * Update ops * 1q gate addition * add gates * Get things closer to working * Get rid of column major * Add Rot gate * Add Rot properly * add to python side * Update parameter access * Update * Apply suggestions from code review Co-authored-by: antalszava <antalszava@gmail.com> Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com> Co-authored-by: antalszava <antalszava@gmail.com> Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com>

* factor out permutations * Remove hadamard and add in operations header * Update ops * 1q gate addition * add gates * Get things closer to working * Get rid of column major * Add Rot gate * Add Rot properly * add to python side * Update parameter access * Update * Create maps for ops * Maps for 1 qubit 3 params; polishing * Extracting the maps out of the functions; declaring them constant; turning to using map.at (constant function) * Moving map defs to header * Remove extern keyword for constants (they are defined too, not just declared) * Moving declarations into if statements * Adding else if so that the second if statement gets executed only if the first was false Co-authored-by: trbromley <brotho02@gmail.com>

* Partially add multiple qubits * Nearly working * Add argsort again * multiqubit version * Remove explicit pass of qubit

* Partially add multiple qubits * Nearly working * Add argsort again * multiqubit version * Adding Rot tests * Remove unnecessary tests * Consts * Perm tests * Argsort tests * Adding basic build.yml * Remove libomp * Installing gtest * Add building and running tests * Correct command format * Running testing with higher permissions * Changing the path * Remove being triggered on push * Adjustments for the integration test suite from PL * Try without sudo * Simplify permutation tests by checking for equality between the expected and the output vectors * Fix a failing test: Co-authored-by: trbromley <brotho02@gmail.com>

* First addition of docs * Update README * Update * Apply suggestions from code review Co-authored-by: antalszava <antalszava@gmail.com> Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com> * Make changes from code review * Add supported ops and obs Co-authored-by: antalszava <antalszava@gmail.com> Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com>

* Partially add multiple qubits * Nearly working * Add argsort again * multiqubit version * Add CRot * Add CRZ * Add CRY * Add target gates * Add CRX * Add CRX * Add phaseshift * Add CZ * add toffoli * Decide not to refactor * Merge better * Add SWAP * Add CSWAP * Don't support MultiRZ * Don't support arbitrary unitaries * Update gitignore * Tests not working * Refactor * Support multiple qubit numbers * Support multiple qubit numbers * Update lightning_qubit.py * Change iterator type * Add CRX etc tests * Add Rot to tests * Remove extra file * Remove extras * Merge properly * Update * Add PhaseShift test * Fix bug * Remove O3 flag * Add Toffoli test * Add CSWAP test * Update qubit number restriction * Remove params from 3q gates: * Apply suggestions from code review * Make Rot tests clearer * Add vectorize() function

* Partially add multiple qubits * Nearly working * Add argsort again * multiqubit version * Add CRot * Add CRZ * Add CRY * Add target gates * Add CRX * Add CRX * Add phaseshift * Add CZ * add toffoli * Decide not to refactor * Merge better * Add SWAP * Add CSWAP * Don't support MultiRZ * Don't support arbitrary unitaries * Update gitignore * Tests not working * Refactor * Support multiple qubit numbers * Support multiple qubit numbers * Update lightning_qubit.py * Change iterator type * Add CRX etc tests * Add Rot to tests * Remove extra file * Remove extras * Merge properly * Update * Add PhaseShift test * Fix bug * Remove O3 flag * Add Toffoli test * Add CSWAP test * Update qubit number restriction * Remove params from 3q gates: * Apply suggestions from code review * Make Rot tests clearer * Add vectorize() function * Update docstrings (#19) * Add docstrings for lightning_qubit.cpp * Update version details * Update init * Minor update * Document lightning_qubit.hpp * Add operations docstrings * Update python docs * Add warning * Apply suggestions from code review Co-authored-by: antalszava <antalszava@gmail.com> Co-authored-by: Josh Izaac <josh146@gmail.com> * Support more gates (#15) * Partially add multiple qubits * Nearly working * Add argsort again * multiqubit version * Add CRot * Add CRZ * Add CRY * Add target gates * Add CRX * Add CRX * Add phaseshift * Add CZ * add toffoli * Decide not to refactor * Merge better * Add SWAP * Add CSWAP * Don't support MultiRZ * Don't support arbitrary unitaries * Update gitignore * Tests not working * Refactor * Support multiple qubit numbers * Support multiple qubit numbers * Update lightning_qubit.py * Change iterator type * Add CRX etc tests * Add Rot to tests * Remove extra file * Remove extras * Merge properly * Update * Add PhaseShift test * Fix bug * Remove O3 flag * Add Toffoli test * Add CSWAP test * Update qubit number restriction * Remove params from 3q gates: * Apply suggestions from code review * Make Rot tests clearer * Add vectorize() function * Apply suggestions from code review Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com> * Apply suggestions from code review * Apply suggestions from code review Co-authored-by: antalszava <antalszava@gmail.com> Co-authored-by: Josh Izaac <josh146@gmail.com> Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com> Co-authored-by: antalszava <antalszava@gmail.com> Co-authored-by: Josh Izaac <josh146@gmail.com> Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com>

* Adding list of supported observables (no Hermitian) * Adding pythontests job * Renaming jobs * Installing PennyLane * Adjusting checking out multiple repos * Correcting path * Installing flaky * Checkout skip_unsupported_ops_tests * Adding the -rsx option to output messages for skipped test cases * Update .github/workflows/build.yml Co-authored-by: Josh Izaac <josh146@gmail.com> * Remove python version miniconda * Removing shots spec for analytic False too * Get PL master and use the --skip-ops option Co-authored-by: Josh Izaac <josh146@gmail.com>

* Partially add multiple qubits * Nearly working * Add argsort again * multiqubit version * Add CRot * Add CRZ * Add CRY * Add target gates * Add CRX * Add CRX * Add phaseshift * Add CZ * add toffoli * Decide not to refactor * Merge better * Add SWAP * Add CSWAP * Don't support MultiRZ * Don't support arbitrary unitaries * Update gitignore * Tests not working * Refactor * Support multiple qubit numbers * Support multiple qubit numbers * Update lightning_qubit.py * Change iterator type * Add CRX etc tests * Add Rot to tests * Remove extra file * Remove extras * Merge properly * Update * Add PhaseShift test * Fix bug * Remove O3 flag * Add Toffoli test * Add CSWAP test * Update qubit number restriction * Remove params from 3q gates: * Apply suggestions from code review * Passing the states and other pars by const ref * Making calc_perm to be an in-place function (by using pass by reference) * Make Rot tests clearer * Add vectorize() function * Reverting to using a tensor_contracted helper variable * Removing the operation parameter when calling contract_3q_op * Update 1q version * Update docstrings (#19) * Add docstrings for lightning_qubit.cpp * Update version details * Update init * Minor update * Document lightning_qubit.hpp * Add operations docstrings * Update python docs * Add warning * Apply suggestions from code review Co-authored-by: antalszava <antalszava@gmail.com> Co-authored-by: Josh Izaac <josh146@gmail.com> * Support more gates (#15) * Partially add multiple qubits * Nearly working * Add argsort again * multiqubit version * Add CRot * Add CRZ * Add CRY * Add target gates * Add CRX * Add CRX * Add phaseshift * Add CZ * add toffoli * Decide not to refactor * Merge better * Add SWAP * Add CSWAP * Don't support MultiRZ * Don't support arbitrary unitaries * Update gitignore * Tests not working * Refactor * Support multiple qubit numbers * Support multiple qubit numbers * Update lightning_qubit.py * Change iterator type * Add CRX etc tests * Add Rot to tests * Remove extra file * Remove extras * Merge properly * Update * Add PhaseShift test * Fix bug * Remove O3 flag * Add Toffoli test * Add CSWAP test * Update qubit number restriction * Remove params from 3q gates: * Apply suggestions from code review * Make Rot tests clearer * Add vectorize() function * Apply suggestions from code review Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com> * Apply suggestions from code review * Apply suggestions from code review Co-authored-by: antalszava <antalszava@gmail.com> Co-authored-by: Josh Izaac <josh146@gmail.com> Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com> Co-authored-by: trbromley <brotho02@gmail.com> Co-authored-by: Tom Bromley <49409390+trbromley@users.noreply.github.com> Co-authored-by: Josh Izaac <josh146@gmail.com> Co-authored-by: Nathan Killoran <co9olguy@users.noreply.github.com>

* Not working * Nearly there * Not there * Closer * Remove shuffling * Remove print * Remove comments * Remove extra letter * Improve docstrings * Rename function and fix tests * Update argsort function * Simplify 1q case * Update 1q case * Update 2q case * Sort out ints in for loops

* Update README * Update requirements * Add to changelog * Remove travis * More tidying * Remove unused tests * Remove qiskit reference * Use github master branch of PL for now * Use pl v0.10 * Try this * Revert to previous requires * Use old req and setup * Replace old test * Lighten setup.py requirements * Update pl v * Update requirements

* Configure setup.py to build the C++ source code * fix * fix * fix * set up CI * typo * typo * more * more * more * more * more * more * more * more * more * more * more * more * more * more * more * please work * fix * fix * fix * Try mac wheels * fix * typo * typo * fix * fix * fix * fix * fix * try something new * try something new * try something new * please work * please work * please work * please work * please work * please work * try again * try again * try again * try again * try again * try again * FINI * remove win32 builds * tweak * Remove manual make file * Update setup.py Co-authored-by: Josh Izaac <josh146@gmail.com> Co-authored-by: trbromley <brotho02@gmail.com> Co-authored-by: Tom Bromley <49409390+trbromley@users.noreply.github.com>

* Update README * Update requirements * Add to changelog * Remove travis * More tidying * Remove unused tests * Remove qiskit reference * Use github master branch of PL for now * Use pl v0.10 * Try this * Revert to previous requires * Use old req and setup * Replace old test * Lighten setup.py requirements * Update pl v * Update requirements * Clearing up the test file for the apply of lightning.qubit; update cpp * Remove unused parts of conftest.py * Add step for unit tests * Update path * Adding pre-def matrices, tol * Adding tests for expval * Adding tests for var * Use internal device object in fixtures * Use internal device object in fixtures more * Update pennylane_lightning/src/tests/lightning_unittest.cpp Co-authored-by: Tom Bromley <49409390+trbromley@users.noreply.github.com> * Update module docstring * Import U from conftest * Copyrights and test file docstrings * Update pennylane_lightning/tests/test_apply.py Co-authored-by: Tom Bromley <49409390+trbromley@users.noreply.github.com> * Add Toffoli test * Remove generate_samples calls * Using device fixture Co-authored-by: trbromley <brotho02@gmail.com> Co-authored-by: Tom Bromley <49409390+trbromley@users.noreply.github.com>

* Update README * Update requirements * Add to changelog * Remove travis * More tidying * Remove unused tests * Remove qiskit reference * Use github master branch of PL for now * Use pl v0.10 * Try this * Revert to previous requires * Use old req and setup * Replace old test * Lighten setup.py requirements * Update pl v * Update requirements * Clearing up the test file for the apply of lightning.qubit; update cpp * Remove unused parts of conftest.py * Add step for unit tests * Update path * Adding pre-def matrices, tol * Adding tests for expval * Adding tests for var * Draft Makefile * Omit tests folder * Adding the test-cpp option * Tweaking dirs * Remove PLUGIN_TESTRUNNER var * tweak test args * Adding google test dir path * Fixing test-cpp * Removing generating samples from var tests too Co-authored-by: trbromley <brotho02@gmail.com>

* Add workflow dispatch event triggering * Add workflow dispatch event triggering (wheels) * Remove newline * Revert to trigger workflows for Testing on pull request

* Add pytest and pytest-cov to requirements * Add 1q test * Add 2q circuit * Add 4 qubit test * Add n-qubit circuit * Tidy * Apply black * Update based on change in PL * Update based on changes in PL * Update due to wires refactor * Quick fix

* Add to README * Mention C++ tests

* First attempt * Update conf * Update conf * Add check in setup.py to not install if USING_RTD environment variable is true * Add an RTD setup file * Change environment variable name * Add install location * Update install location * Update install location * Update * Add install * Update setup.py * Update requirements

* Have new path in workflow file * Updating remaining occurrences

* Use Windows optimization compiler flag * Further modifying compiler and linker options * Specifying C++14 option * Remove C++ version pinning * Sepcify C++14 version comp option * Specifying the -std:c++11 option and turning to using the - character * Building the wheel on push and on pull request * Correct event trigger Co-authored-by: trbromley <brotho02@gmail.com>

* Tidy up * Extra update * Update setup.py Co-authored-by: antalszava <antalszava@gmail.com> * Apply suggestion Co-authored-by: antalszava <antalszava@gmail.com>

* Add black check * Run black * Add codecov report generation * Add codecov report generation * Update location * Update location * Update for correct path * Increase coverage * Isort * Add shields * Sort out code quality * Update build action * Apply suggestions from code review Co-authored-by: antalszava <antalszava@gmail.com> * Use pytest mock Co-authored-by: antalszava <antalszava@gmail.com>

* Incrementing the version number & requirement number * fixes * fix * fix Co-authored-by: Josh Izaac <josh146@gmail.com>

* Turning to using the GitLab files for Eigen * Fixes

AmintorDusko · 2022-09-22T12:41:06Z

Hi @multiphaseCFD, nice work so far!
Can we have unit tests for the new methods?
I wonder why in some of your benchmarks the methods executed with float precision are taking longer than their double precision counterparts. Do you have any insight about that?

multiphaseCFD · 2022-09-22T13:38:18Z

Hi @multiphaseCFD, nice work so far! Can we have unit tests for the new methods? I wonder why in some of your benchmarks the methods executed with float precision are taking longer than their double precision counterparts. Do you have any insight about that?

Thanks @AmintorDusko ! In fact, thanks to @chaeyeunpark 's nice implementation, the newly implemented gates are implicitly tested by the existing testing kernels here. Do you know how to pass-by this errors according to the current situation?
Good question for why double is faster. I believe it's related to hardware architecture. According to tests on Mac M1, fp32 computation is faster than fp64 computation as shown below.

AmintorDusko · 2022-09-22T13:49:26Z

Hi @multiphaseCFD, nice work so far! Can we have unit tests for the new methods? I wonder why in some of your benchmarks the methods executed with float precision are taking longer than their double precision counterparts. Do you have any insight about that?

Thanks @AmintorDusko ! In fact, thanks to @chaeyeunpark 's nice implementation, the newly implemented gates are implicitly tested by the existing testing kernels here. Do you know how to pass-by this errors according to the current situation? Good question for why double is faster. I believe it's related to hardware architecture. According to tests on Mac M1, fp32 computation is faster than fp64 computation as shown below.

Thank you, @multiphaseCFD, for addressing my questions.

chaeyeunpark

Thanks @multiphaseCFD for amazing work! I haven't expected to have LM implementations of four qubit gates (are they in Kokkos?). Let's see the coverage report after merging #359. If it looks fine, I will approve this.

pennylane_lightning/src/gates/cpu_kernels/GateImplementationsLM.hpp

chaeyeunpark · 2022-09-22T16:55:39Z

I also note that we can let StateVector classes use these kernels by giving information in https://github.com/PennyLaneAI/pennylane-lightning/blob/master/pennylane_lightning/src/simulator/AssignKernelMap_Default.cpp. We may have that as a separate PR.

multiphaseCFD · 2022-09-22T18:20:13Z

Thanks @multiphaseCFD for amazing work! I haven't expected to have LM implementations of four qubit gates (are they in Kokkos?). Let's see the coverage report after merging #359. If it looks fine, I will approve this.

Thanks @chaeyeunpark ! Yes, I implemented them in L-Kokkos.

* split coverage and unit tests for cpptests * spread changes to all Linux cpptests

* set wheel build matrix in a reusable workflow (linux) * input event_name to set wheel build script * propagate changes to all wheels

chaeyeunpark

Thanks @multiphaseCFD ! One but non-essential comment.

pennylane_lightning/src/benchmarks/Bench_LinearAlgebra.cpp

mlxd

Just a quick one from me

mlxd · 2022-09-23T19:17:04Z

pennylane_lightning/src/gates/cpu_kernels/GateImplementationsLM.hpp

@@ -66,6 +66,54 @@ class GateImplementationsLM : public PauliGenerator<GateImplementationsLM> {
        return {parity_high, parity_middle, parity_low};
    }

+    static auto revWireParity(size_t rev_wire0, size_t rev_wire1,


We can probably define this and the next one as inline. Also, worth a try to make it constexpr too, (need to check if fillTrailingOnes and fillLeadingOnes are compatible).

Also, since we know the return type, it would be good to add it as an arrow type: -> std::array<size_t,parity_size>

mlxd · 2022-09-26T12:58:18Z

@multiphaseCFD Can we not just reopen this PR and keep everything here? I'd rather avoid splitting the reviews and details over 2 separate PR processes.

antalszava and others added 30 commits July 22, 2020 11:05

Merge branch 'master' into add_skeleton

5c151d8

Merge pull request #5 from XanaduAI/add_skeleton

58aef73

Adds skeleton for C++ section of lightning.qubit

Extend to different qubit numbers (#12)

060b249

* Partially add multiple qubits * Nearly working * Add argsort again * multiqubit version * Remove explicit pass of qubit

Apply code formatter (#24)

93fc0ea

Change to triggering CI jobs manually (#30)

2b8a0ea

* Add workflow dispatch event triggering * Add workflow dispatch event triggering (wheels) * Remove newline * Revert to trigger workflows for Testing on pull request

Benchmarking and polishing README (#27)

d3fd440

* Add to README * Mention C++ tests

Move the tests folder one level up (Python) (#35)

52c93f1

* Have new path in workflow file * Updating remaining occurrences

Tidy up (#36)

bbf903c

* Tidy up * Extra update * Update setup.py Co-authored-by: antalszava <antalszava@gmail.com> * Apply suggestion Co-authored-by: antalszava <antalszava@gmail.com>

Incrementing the version number to 0.11.0 (#40)

5e2ed3f

* Incrementing the version number & requirement number * fixes * fix * fix Co-authored-by: Josh Izaac <josh146@gmail.com>

Increment version number to v0.12 (#41)

8b0da65

Update devices.rst

a9013c5

Eigen from GitLab (#44)

8911379

* Turning to using the GitLab files for Eigen * Fixes

chaeyeunpark reviewed Sep 22, 2022

View reviewed changes

pennylane_lightning/src/gates/cpu_kernels/GateImplementationsLM.hpp Outdated Show resolved Hide resolved

Shuli Shu added 2 commits September 22, 2022 17:11

use std::sort to sort wires

461c971

return type std::tuple->std::array

4a2d0c8

AmintorDusko and others added 11 commits September 22, 2022 14:49

Split coverage and unit tests (#359)

5022b84

* split coverage and unit tests for cpptests * spread changes to all Linux cpptests

Merge branch 'master' into lm_impl_3wires_gate

f42f789

merge main & initilize parity with zeros

385fd32

Auto update version

8454431

Tidy (C++) test

ccadc80

Reusable workflow for wheel build matrix (#357)

2f88a1a

* set wheel build matrix in a reusable workflow (linux) * input event_name to set wheel build script * propagate changes to all wheels

Tidy C++ fix

45eb6eb

Merge branch 'master' into lm_impl_3wires_gate

e182661

Auto update version

e786178

make format

534c749

remove commented lines

faa6ad5

multiphaseCFD requested review from chaeyeunpark, mlxd and AmintorDusko September 22, 2022 21:26

chaeyeunpark approved these changes Sep 22, 2022

View reviewed changes

pennylane_lightning/src/benchmarks/Bench_LinearAlgebra.cpp Outdated Show resolved Hide resolved

pennylane_lightning/src/benchmarks/Bench_LinearAlgebra.cpp Outdated Show resolved Hide resolved

update Bench_LinearAlgebra

8353493

mlxd reviewed Sep 23, 2022

View reviewed changes

multiphaseCFD closed this Sep 23, 2022

multiphaseCFD force-pushed the lm_impl_3wires_gate branch from 97a2bc4 to 8353493 Compare September 23, 2022 21:40

multiphaseCFD added WIP Work-in-progress issue. Not for general fixing. and removed WIP Work-in-progress issue. Not for general fixing. labels Sep 26, 2022

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

LM implementation of multi-wire gates #358

LM implementation of multi-wire gates #358

multiphaseCFD commented Sep 21, 2022 •

edited

AmintorDusko commented Sep 22, 2022 •

edited

multiphaseCFD commented Sep 22, 2022

AmintorDusko commented Sep 22, 2022

chaeyeunpark left a comment

chaeyeunpark commented Sep 22, 2022

multiphaseCFD commented Sep 22, 2022

chaeyeunpark left a comment

mlxd left a comment

mlxd Sep 23, 2022

multiphaseCFD Sep 23, 2022

mlxd commented Sep 26, 2022

LM implementation of multi-wire gates #358

LM implementation of multi-wire gates #358

Conversation

multiphaseCFD commented Sep 21, 2022 • edited

Before submitting

AmintorDusko commented Sep 22, 2022 • edited

multiphaseCFD commented Sep 22, 2022

AmintorDusko commented Sep 22, 2022

chaeyeunpark left a comment

Choose a reason for hiding this comment

chaeyeunpark commented Sep 22, 2022

multiphaseCFD commented Sep 22, 2022

chaeyeunpark left a comment

Choose a reason for hiding this comment

mlxd left a comment

Choose a reason for hiding this comment

mlxd Sep 23, 2022

Choose a reason for hiding this comment

multiphaseCFD Sep 23, 2022

Choose a reason for hiding this comment

mlxd commented Sep 26, 2022

multiphaseCFD commented Sep 21, 2022 •

edited

AmintorDusko commented Sep 22, 2022 •

edited