Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
Menu
Open sidebar
OpenDAS
DeepEP
Commits
d24bbeba
Commit
d24bbeba
authored
Jun 27, 2025
by
alpha-baby
Browse files
opt code
parent
b90320e2
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
1 addition
and
1 deletion
+1
-1
csrc/kernels/internode.cu
csrc/kernels/internode.cu
+1
-1
No files found.
csrc/kernels/internode.cu
View file @
d24bbeba
...
@@ -125,7 +125,7 @@ notify_dispatch(const int* num_tokens_per_rank, int* moe_recv_counter_mapped, in
...
@@ -125,7 +125,7 @@ notify_dispatch(const int* num_tokens_per_rank, int* moe_recv_counter_mapped, in
// Issue send
// Issue send
// TODO: more light fence or barrier or signaling
// TODO: more light fence or barrier or signaling
// TODO: overlap EP barrier and NVL cleaning
// TODO: overlap EP barrier and NVL cleaning
for
(
int
i
=
warp_id
;
i
<
kNumRDMARanks
;
i
+=
num_
threads
/
32
)
{
for
(
int
i
=
warp_id
;
i
<
kNumRDMARanks
;
i
+=
num_
warps
)
{
if
(
i
!=
rdma_rank
)
{
if
(
i
!=
rdma_rank
)
{
nvshmemi_ibgda_put_nbi_warp
<
true
>
(
reinterpret_cast
<
uint64_t
>
(
rdma_recv_num_tokens_mixed
.
recv_buffer
(
rdma_rank
)),
nvshmemi_ibgda_put_nbi_warp
<
true
>
(
reinterpret_cast
<
uint64_t
>
(
rdma_recv_num_tokens_mixed
.
recv_buffer
(
rdma_rank
)),
reinterpret_cast
<
uint64_t
>
(
rdma_recv_num_tokens_mixed
.
send_buffer
(
i
)),
reinterpret_cast
<
uint64_t
>
(
rdma_recv_num_tokens_mixed
.
send_buffer
(
i
)),
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment