Runtime Parameters
This page is generated from:
src/g4integration/config/AdePTConfigurationMessenger.cc(authoritative command definitions and guidance)test/regression/scripts/test_ui_commands_template.mac(example invocations used in CI)
Regenerate with:
python3 docs/scripts/generate_runtime_parameters.py
Categories are sourced from // ADEPT_DOCS_SECTION: ... markers in
src/g4integration/config/AdePTConfigurationMessenger.cc.
Examples come from the CI UI command macro. If a command shows - in the
Example column, add an invocation in test_ui_commands_template.mac.
Misc
Command |
Value type |
Description |
Constraints |
Example |
|---|---|---|---|---|
|
integer |
Set verbosity level for the AdePT integration layer |
- |
|
|
integer |
Set the base seed for the rng. Default: 1234567 |
- |
|
Setting Up the GPU
Command |
Value type |
Description |
Constraints |
Example |
|---|---|---|---|---|
|
double |
Set the total number of track slots that will be allocated on the GPU, in millions |
- |
|
|
double |
Set the total number of hit slots that will be allocated on the GPU, in millions |
- |
|
|
double |
Set the usage threshold at which the GPU steps are copied from the buffer and not taken directly by the G4 workers |
parameter |
|
|
double |
Sets the CPUCapacity factor for returned GPU steps with respect to the GPU (see: /adept/setMillionsOfHitSlots). Must at least be 2.5 |
parameter |
|
|
double |
Sets the HitBuffer safety factor for stalling the GPU. If nParticlesInFlight * HitBufferSafetyFactor > NumHitSlotsLeft, the GPU will stall. Default: 1.5 |
- |
|
|
integer |
Set the CUDA device stack limit |
- |
|
|
integer |
Set the CUDA device heap limit |
- |
|
Specify the regions where the GPU is used
Command |
Value type |
Description |
Constraints |
Example |
|---|---|---|---|---|
|
bool |
If true, particles are tracked on the GPU across the whole geometry |
- |
|
|
string |
Add a region in which transport will be done on GPU |
- |
|
|
string |
Remove a region in which transport will be done on GPU (so it will be done on the CPU) |
- |
|
User Actions
Command |
Value type |
Description |
Constraints |
Example |
|---|---|---|---|---|
|
bool |
If true, the UserSteppingAction is called for on every step. WARNING: The steps are currently not sorted, that means it is not guaranteed that the UserSteppingAction is called in order, i.e., it could get called on the secondary before the primary has finished its track. NOTE: This means that every single step is recorded on GPU and send back to CPU, which can impact performance |
- |
|
|
bool |
If true, the PostUserTrackingAction is called for on every track. NOTE: This means that the last step of every track is recorded on GPU and send back to CPU |
- |
|
Special Settings
Command |
Value type |
Description |
Constraints |
Example |
|---|---|---|---|---|
|
bool |
If true, all electrons, positrons, gammas handed over to AdePT are immediately killed. WARNING: Only to be used for testing the speed and fraction of EM, all results are wrong! |
- |
|
|
string |
Temporary method for setting the geometry to use with VecGeom |
- |
|
|
string |
Set the path the the covfie file for reading in an external magnetic field |
- |
- |
|
integer |
Set N, the number of last N particles per event that are finished on CPU. Default: 0. This is an important parameter for handling loopers in a magnetic field |
- |
|
|
integer |
Set N, the number of maximum Woodcock tracking iterations per step before giving the gamma back to the normal gamma kernel. Default: 5. This can be used to optimize the performance in highly granular geometries |
- |
|
Special settings for G4EmStandard_AdePT physics constructor
Command |
Value type |
Description |
Constraints |
Example |
|---|---|---|---|---|
|
string |
Add a region in which the gamma transport is done via Woodcock tracking. NOTE: This ONLY applies to the AdePTPhysics, if the PhysicsList uses ANY other physics (which is done in LHCb, CMS, ATLAS) then this will have NO effect! |
- |
|
|
double |
Sets a kinetic energy limit above which the gamma transport is done via Woodcock tracking in the assigned regions. NOTE: This ONLY applies to the AdePTPhysics, if the PhysicsList uses ANY other physics (which is done in LHCb, CMS, ATLAS) then this will have NO effect! |
- |
|
|
bool |
If true, this configures G4HepEm to use multiple steps in MSC on CPU. This does not affect GPU transport |
- |
|
|
bool |
If true, this configures G4HepEm to use energy loss fluctuations. This affects both CPU and GPU transport NOTE: This is only true for the AdePTPhysics in the examples! In all other physics lists the setting is taken directly from Geant4 and this parameter does not change it. |
- |
|