Commit 9904929b authored by Jing Zhang's avatar Jing Zhang
Browse files

fixed splitK

parent e731e8c0
...@@ -95,7 +95,7 @@ struct DeviceGemmXdlSplitKCShuffle : public DeviceGemmSplitK<ALayout, ...@@ -95,7 +95,7 @@ struct DeviceGemmXdlSplitKCShuffle : public DeviceGemmSplitK<ALayout,
const auto a_grid_desc_m_kpad = transform_tensor_descriptor( const auto a_grid_desc_m_kpad = transform_tensor_descriptor(
a_grid_desc_m_k, a_grid_desc_m_k,
make_tuple(make_right_pad_transform(K, KPad - K), make_pass_through_transform(M)), make_tuple(make_pass_through_transform(M), make_right_pad_transform(K, KPad - K)),
make_tuple(Sequence<0>{}, Sequence<1>{}), make_tuple(Sequence<0>{}, Sequence<1>{}),
make_tuple(Sequence<0>{}, Sequence<1>{})); make_tuple(Sequence<0>{}, Sequence<1>{}));
......
...@@ -547,11 +547,6 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2 ...@@ -547,11 +547,6 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
static_cast<FloatC*>(p_shared_block), static_cast<FloatC*>(p_shared_block),
c_block_desc_mblock_mperblock_nblock_nperblock.GetElementSpaceSize()); c_block_desc_mblock_mperblock_nblock_nperblock.GetElementSpaceSize());
static_assert(M1 == MWave, "");
static_assert(N1 == NWave, "");
static_assert(M2 * M3 * M4 == MPerXDL, "");
static_assert(N2 == NPerXDL, "");
constexpr auto c_block_desc_m0_n0_m1_n1_m2_m3_m4_n2 = transform_tensor_descriptor( constexpr auto c_block_desc_m0_n0_m1_n1_m2_m3_m4_n2 = transform_tensor_descriptor(
c_block_desc_mblock_mperblock_nblock_nperblock, c_block_desc_mblock_mperblock_nblock_nperblock,
make_tuple( make_tuple(
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment