GPUOcelot
|
#include <OcelotConfiguration.h>
Public Member Functions | |
Executive () | |
Public Attributes | |
int | defaultDeviceID |
ID of device to select by default. | |
int | preferredISA |
selects preferred ISA [if multiple are available]
| |
int | optimizationLevel |
selects optimization level - cast to translator::Translator::OptimizationLevel | |
bool | required |
if true, Ocelot will either choose only | |
bool | enableLLVM |
filter enabling LLVM devices if present | |
bool | enableEmulated |
filter enbabling Emulated PTX devices if present | |
bool | enableNVIDIA |
filter enabling CUDA GPU devices if present | |
bool | enableAMD |
filter enabling AMD GPU devices if present | |
bool | enableRemote |
filter enabling remote GPU devices if present | |
int | port |
The port to connect to an ocelot server for remoting. | |
std::string | host |
The hostname of the ocelot server to connect. | |
int | reconvergenceMechanism |
Reconvergence mechanism. | |
int | workerThreadLimit |
maximum number of worker threads - a non-positive | |
int | warpSize |
suggested warp size |
api::OcelotConfiguration::Executive::Executive | ( | ) |
ID of device to select by default.
filter enabling AMD GPU devices if present
filter enbabling Emulated PTX devices if present
filter enabling LLVM devices if present
filter enabling CUDA GPU devices if present
filter enabling remote GPU devices if present
std::string api::OcelotConfiguration::Executive::host |
The hostname of the ocelot server to connect.
selects optimization level - cast to translator::Translator::OptimizationLevel
The port to connect to an ocelot server for remoting.
selects preferred ISA [if multiple are available]
Reconvergence mechanism.
if true, Ocelot will either choose only
suggested warp size
maximum number of worker threads - a non-positive