issue/949: add silu_and_mul for moore gpu with test pass#970

Closed

spike-zhu wants to merge 31 commits intomainfrom

Commits on Jan 27, 2026

issue/987 - add .cpp files to ninetoothed includes
wooway777
committed
issue/978 - metax cuda graph impl and wrappings
wooway777
committed
issue/846 - Refactor embedding to support device-side input and CUDA graph recording

gongchensu
authored and
wooway777
committed
issue/900 - support embedding on iluvatar, metax, and moore
wooway777
committed
issue/900 - adapt to graph and adjust test script
wooway777
committed
issue/900 - maintains classic embedding for devices yet to be worked on
wooway777
committed
issue/791 fix add_rmsnorm api and rmsnorm module

PanZezhong1725
authored and
wooway777
committed
issue/884 - add_rms_norm on iluvatar, metax and moore
wooway777
committed
issue/632 - adapt to iluvatar core 20
wooway777
committed
issue/791 - fix add_rmsnorm api on mtx and mth
wooway777
committed
issue/810 support more ops as graph op

PanZezhong1725
authored and
wooway777
committed
issue/985 - adjust cxflags and cxxflags for lua scripts
wooway777
committed
issue/402 - convenient ninetoothed util

voltjia
authored and
wooway777
committed
issue/925 - Speed up `scripts/build_ntops.py` and `src/infiniop/ninetoothed/build.py` with `concurrent.futures`

voltjia
authored and
wooway777
committed
issue/940 - check build result and implicitly require build.py for build ntops
wooway777
committed
issue/935 - add metax include dir for ninetoothed
wooway777
committed
issue/919 - ninetoothed flash attention
wooway777
committed
issue/931 - ninetoothed swiglu for nv, il, mtx
wooway777
committed
issue/923 - ninetoothed kv caching for nv, il, mtx
wooway777
committed
issue/979 optimize paged attention

PanZezhong1725
authored and
wooway777
committed
issue/979 - removed commented paged attn codes
wooway777
committed
issue/983 - adapted the optimized paged attention to metax
wooway777
committed
demo131 - patch lua flags and includes
wooway777
committed
issue/811 use relax graph capture mode, add compile flag for graph instantiate
PanZezhong1725
committed
Merge pull request #989 from InfiniTensor/issue/811-fix
PanZezhong1725
authored

Commits on Jan 29, 2026

issue/995 fix paged attn on iluvatar
zhangyue207
authored

Commits on Feb 10, 2026

issue/949 - feat: add silu_and_mul for moore gpu with test pass
spike-zhu
committed