Triton-distributed-ascend
Language
English
中文
Getting Started
Installation
Build from Source
Quick Start
Release Note
Architecture
Distributed Architecture Design
API Reference
Triton-distributed Semantics
triton_dist.language API
SHMEM Device API
SHMEM Host API
Swizzle Helper Functions
Autotuning API
Developer Guide
Kernel Debugging
Environment Variables and Compiler Options
Multi-Card Communication-Computation Fusion Operator Precision Verification
Single Operator Development
Performance Optimization Guide
Operator Performance Testing and Tuning: autotune Feature Usage Guide
Unified Ascend operator tests
Tutorials
AllGather-GEMM (Allgather Gemm)
Autotuning Demo: UDMA Reverse All2All Optimization
Reverse All2All (EP MoE Distributed Kernel)
FAQ
FAQ Documentation
Triton-distributed-ascend
Index
Index
A
|
B
|
C
|
E
|
F
|
G
|
I
|
L
|
M
|
N
|
O
|
P
|
Q
|
R
|
S
|
T
|
W
A
aclshmem_align()
built-in function
aclshmem_calloc()
built-in function
aclshmem_create_tensor()
built-in function
aclshmem_finalize()
built-in function
aclshmem_free()
built-in function
aclshmem_free_tensor()
built-in function
aclshmem_get_unique_id()
built-in function
aclshmem_getmem()
built-in function
aclshmem_getmem_nbi()
built-in function
aclshmem_global_exit()
built-in function
aclshmem_info_get_name()
built-in function
aclshmem_info_get_version()
built-in function
aclshmem_init()
built-in function
aclshmem_init_using_unique_id()
built-in function
aclshmem_malloc()
built-in function
aclshmem_ptr()
built-in function
aclshmem_putmem()
built-in function
aclshmem_putmem_nbi()
built-in function
aclshmem_signal_wait_until()
built-in function
ACLSHMEMCmpOp (built-in class)
ACLSHMEMSignalOp (built-in class)
ACLSHMEMTeam (built-in class)
aclshmemx_get_heap_base()
built-in function
aclshmemx_putmem_signal()
built-in function
aclshmemx_putmem_signal_nbi()
built-in function
ADD (ACLSHMEMSignalOp attribute)
B
barrier()
built-in function
barrier_all()
built-in function
barrier_all_vec()
built-in function
barrier_vec()
built-in function
built-in function
aclshmem_align()
aclshmem_calloc()
aclshmem_create_tensor()
aclshmem_finalize()
aclshmem_free()
aclshmem_free_tensor()
aclshmem_get_unique_id()
aclshmem_getmem()
aclshmem_getmem_nbi()
aclshmem_global_exit()
aclshmem_info_get_name()
aclshmem_info_get_version()
aclshmem_init()
aclshmem_init_using_unique_id()
aclshmem_malloc()
aclshmem_ptr()
aclshmem_putmem()
aclshmem_putmem_nbi()
aclshmem_signal_wait_until()
aclshmemx_get_heap_base()
aclshmemx_putmem_signal()
aclshmemx_putmem_signal_nbi()
barrier()
barrier_all()
barrier_all_vec()
barrier_vec()
consume_token()
extern_call()
fence()
getmem()
getmem_nbi()
int_p()
my_pe()
n_pes()
notify()
num_ranks()
pe_count()
putmem()
putmem_nbi()
putmem_signal()
putmem_signal_nbi()
quiet()
rank()
remote_ptr()
set_log_level()
signal_op()
signal_wait_until()
symm_at()
team_destroy()
team_my_pe()
team_n_pes()
team_split_2d()
team_split_strided()
team_translate_pe()
wait()
C
consume_token()
built-in function
E
EQ (ACLSHMEMCmpOp attribute)
extern_call()
built-in function
F
fence()
built-in function
G
GE (ACLSHMEMCmpOp attribute)
getmem()
built-in function
getmem_nbi()
built-in function
GT (ACLSHMEMCmpOp attribute)
I
InitAttr (built-in class)
InitStatus (built-in class)
int_p()
built-in function
INVALID (ACLSHMEMTeam attribute)
L
LE (ACLSHMEMCmpOp attribute)
LT (ACLSHMEMCmpOp attribute)
M
mem_size (InitAttr attribute)
my_pe()
built-in function
N
n_pes()
built-in function
NE (ACLSHMEMCmpOp attribute)
notify()
built-in function
num_ranks()
built-in function
O
OpEngineType (built-in class)
P
pe_count()
built-in function
putmem()
built-in function
putmem_nbi()
built-in function
putmem_signal()
built-in function
putmem_signal_nbi()
built-in function
Q
quiet()
built-in function
R
rank()
built-in function
remote_ptr()
built-in function
S
SET (ACLSHMEMSignalOp attribute)
set_log_level()
built-in function
signal_op()
built-in function
signal_wait_until()
built-in function
symm_at()
built-in function
T
team_destroy()
built-in function
team_my_pe()
built-in function
team_n_pes()
built-in function
team_split_2d()
built-in function
team_split_strided()
built-in function
team_translate_pe()
built-in function
W
wait()
built-in function
WORLD (ACLSHMEMTeam attribute)