noc_async_read_shard

template<typename DSpec>
void noc_async_read_shard(const uint32_t shard_id, const TensorAccessor<DSpec> &s, std::uint32_t dst_local_l1_addr, uint8_t noc = noc_index)

Initiates an asynchronous read of a shard from a source noc address into a local L1 address. The size of the transaction and the source address are determined by the TensorAccessor object. This function only works for sharded tensors.

Return value: None

Argument	Description	Type	Valid Range	Required
shard_id	Row-major index of a shard in the sharded tensor	uint32_t	Any uint32_t number	True
s	TensorAccessor object	TensorAccessor	Any TensorAccessor object, refer to tensor_accessor.h	True
dst_local_l1_addr	Destination address in local L1 memory	uint32_t	0..1MB	True
noc	Which NOC to use for the transaction	uint8_t	0 or 1	False