Memory¶
memory.param¶
-
namespace
picongpu
-
Typedefs
-
typedef mCT::shrinkTo<mCT::Int<8, 8, 4>, simDim>::type picongpu
SuperCellSize
size of a superCell
volume of a superCell must be <= 1024
-
typedef MappingDescription<simDim, SuperCellSize> picongpu
MappingDesc
define mapper which is used for kernel call mappings
Variables
-
constexpr size_t picongpu
reservedGpuMemorySize
= 350 *1024*1024
-
constexpr uint32_t picongpu
GUARD_SIZE
= 1
-
constexpr uint32_t picongpu
BYTES_EXCHANGE_X
= 4 * 256 * 1024 how many bytes for buffer is reserved to communication in one direction
-
constexpr uint32_t picongpu
BYTES_EXCHANGE_Y
= 6 * 512 * 1024
-
constexpr uint32_t picongpu
BYTES_EXCHANGE_Z
= 4 * 256 * 1024
-
constexpr uint32_t picongpu
BYTES_CORNER
= 8 * 1024
-
constexpr uint32_t picongpu
BYTES_EDGES
= 32 * 1024
-
constexpr uint32_t picongpu
fieldTmpNumSlots
= 1 number of scalar fields that are reserved as temporary fields
-
constexpr bool picongpu
fieldTmpSupportGatherCommunication
= true can
FieldTmp
gather neighbor informationIf
true
it is possible to call the methodasyncCommunicationGather()
to copy data from the border of neighboring GPU into the local guard. This is also known as building up a “ghost” or “halo” region in domain decomposition and only necessary for specific algorithms that extend the basic PIC cycle, e.g. with dependence on derived density or energy fields.
-
typedef mCT::shrinkTo<mCT::Int<8, 8, 4>, simDim>::type picongpu
precision.param¶
-
namespace
picongpu
mallocMC.param¶
-
namespace
picongpu
-
Typedefs
-
using
picongpu::DeviceHeap = typedef mallocMC::Allocator< mallocMC::CreationPolicies::Scatter< DeviceHeapConfig >, mallocMC::DistributionPolicies::Noop, mallocMC::OOMPolicies::ReturnNull, mallocMC::ReservePoolPolicies::SimpleCudaMalloc, mallocMC::AlignmentPolicies::Shrink<> >
-
struct picongpu
DeviceHeapConfig
Public Types
-
using picongpu::DeviceHeapConfig
pagesize
= boost::mpl::int_<2 * 1024 * 1024>
-
using picongpu::DeviceHeapConfig
accessblocks
= boost::mpl::int_<4>
-
using picongpu::DeviceHeapConfig
regionsize
= boost::mpl::int_<8>
-
using picongpu::DeviceHeapConfig
wastefactor
= boost::mpl::int_<2>
-
using picongpu::DeviceHeapConfig
resetfreedpages
= boost::mpl::bool_<true>
-
using picongpu::DeviceHeapConfig
-
using