|
namespace | executor_recycling |
| CPPuddle namespace containing the executor pool functionality.
|
|
namespace | kernel_aggregation |
| CPPuddle namespace containing the kernel aggregation functionality.
|
|
namespace | memory_recycling |
| CPPuddle namespace containing the memory recycling functionality.
|
|
|
size_t | get_device_id (const size_t number_gpus) |
| Uses HPX thread information to determine which GPU should be used.
|
|
Primary CPPuddle namespace containing the three primary feature modules / (memory_recycling, / executor_recycling and kernel_aggregation) in sub-namespaces
◆ mutex_t
◆ get_device_id()
size_t cppuddle::get_device_id |
( |
const size_t |
number_gpus | ) |
|
|
inline |
Uses HPX thread information to determine which GPU should be used.
◆ max_number_gpus
constexpr size_t cppuddle::max_number_gpus = CPPUDDLE_HAVE_MAX_NUMBER_GPUS |
|
constexpr |
◆ number_instances
constexpr size_t cppuddle::number_instances = CPPUDDLE_HAVE_NUMBER_BUCKETS |
|
constexpr |