HcclCommInitClusterInfoConfig

Applicability

Product	Supported
Atlas 350 Accelerator Card	√
Atlas A3 training product/Atlas A3 inference product	√
Atlas A2 training product/Atlas A2 inference product	√
Atlas 200I/500 A2 inference product	☓
Atlas inference product	√
Atlas training product	√

For Atlas A2 training product/Atlas A2 inference product, only the Atlas 800T A2 training server, Atlas 900 A2 PoD cluster basic unit, and Atlas 200T A2 Box16 heterogeneous subrack are supported.

For the Atlas inference products, only the Atlas 300I Duo inference card is supported.

Function

Initializes HCCL based on the rank table and creates an HCCL communicator with specific configurations.

Prototype

HcclResult HcclCommInitClusterInfoConfig(const char *clusterInfo, uint32_t rank, HcclCommConfig *config, HcclComm *comm)

Parameters

Parameter	Input/Output	Description
clusterInfo	Input	Directory (including the file name) of the rank table file, which is of the string type and contains a maximum of 4096 bytes, including the end character.
rank	Input	ID of the current rank. Note that the value of this parameter must be the same as the value of rank_id in the rank table file.
config	Input	Communicator configuration items, including the buffer size, deterministic computing switch, communicator name, and communication operator expansion mode. Configuration parameters must fall within the valid value range. For details on the parameters and their priorities in HcclCommConfig, see HcclCommConfig. Note that the input config must be first initialized by calling HcclCommConfigInit.
comm	Output	Pointer to the initialized communicator. For details about the definition of the HcclComm type, see HcclComm.

Returns

HcclResult: HCCL_SUCCESS on success, or else failure.

Constraints

Repeated initialization is not supported in the same communicator.

Call Example

// Initialize device resources.
aclInit(NULL);
// Directory of the rank table configuration file
const char *rankTableFile = "/path/to/rank_table.json";
// Device used for the collective communication operations
uint32_t rankSize = 8;
uint32_t devId = 0;
aclrtSetDevice(devId);
// Create and initialize the configuration items of the communicator.
HcclCommConfig config;
HcclCommConfigInit(&config);
// Modify the communicator configuration as required.
config.hcclBufferSize = 50;  // Size of the buffer for storing the shared data, in MB. The value must be greater than or equal to 1. The default value is 200.
std::strcpy(config.hcclCommName, "comm_1");
// Initialize the communicator.
HcclComm hcclComm;
// In this example, devId is used as the rank ID of the current rank.
HcclCommInitClusterInfoConfig(rankTableFile, devId, &config, &hcclComm);
// Destroy the communicator.
HcclCommDestroy(hcclComm);
// Deinitialize device resources.
aclFinalize();

Parent topic: C APIs