leavelet/cutlass
1
0
Fork 0
Code Issues Pull Requests Actions 24 Packages Projects Releases Wiki Activity

Updates for CUTLASS 3.4.1 (#1346)

Browse Source 
* Updates for CUTLASS 3.4.1

* minor epi change
This commit is contained in:
ANIKET SHIVAM
2024-02-15 15:48:34 -05:00
committed by GitHub
parent 47a3ebbea9
commit bbe579a9e3
49 changed files with 800 additions and 451 deletions
Download Patch File Download Diff File Expand all files Collapse all files
include/cutlass/gemm/dispatch_policy.hpp
+4 -7
View File
@@ -53,8 +53,7 @@ struct KernelTma { };
struct KernelTmaWarpSpecialized { };
struct KernelTmaWarpSpecializedPingpong { };
struct KernelTmaWarpSpecializedCooperative { };
struct KernelArrayTmaWarpSpecializedCooperative { };
struct KernelGroupTmaWarpSpecializedCooperative { };
struct KernelPtrArrayTmaWarpSpecializedCooperative { };
//////////////////////////////////////////////////////////////////////////////
@@ -67,8 +66,7 @@ struct KernelGroupTmaWarpSpecializedCooperative { };
struct KernelTmaWarpSpecializedFP8FastAccum : KernelTmaWarpSpecialized { };
struct KernelTmaWarpSpecializedPingpongFP8FastAccum : KernelTmaWarpSpecializedPingpong { };
struct KernelTmaWarpSpecializedCooperativeFP8FastAccum: KernelTmaWarpSpecializedCooperative { };
struct KernelArrayTmaWarpSpecializedCooperativeFP8FastAccum : KernelArrayTmaWarpSpecializedCooperative { };
struct KernelGroupTmaWarpSpecializedCooperativeFP8FastAccum : KernelGroupTmaWarpSpecializedCooperative { };
struct KernelPtrArrayTmaWarpSpecializedCooperativeFP8FastAccum : KernelPtrArrayTmaWarpSpecializedCooperative { };
// Policies to opt into mixed type GEMMs
struct KernelTmaWarpSpecializedMixedInput : KernelTmaWarpSpecialized { };
@@ -233,7 +231,7 @@ struct MainloopSm90TmaGmmaWarpSpecializedFP8
template<
int Stages_,
class ClusterShape_ = Shape<_1,_1,_1>,
class KernelSchedule = KernelGroupTmaWarpSpecializedCooperative
class KernelSchedule = KernelPtrArrayTmaWarpSpecializedCooperative
>
struct MainloopSm90ArrayTmaGmmaWarpSpecialized {
constexpr static int Stages = Stages_;
@@ -241,8 +239,7 @@ struct MainloopSm90ArrayTmaGmmaWarpSpecialized {
using ArchTag = arch::Sm90;
using Schedule = KernelSchedule;
static_assert(
cute::is_base_of_v<KernelArrayTmaWarpSpecializedCooperative, KernelSchedule> ||
cute::is_base_of_v<KernelGroupTmaWarpSpecializedCooperative, KernelSchedule>,
cute::is_base_of_v<KernelPtrArrayTmaWarpSpecializedCooperative, KernelSchedule>,
"KernelSchedule must be one of the Ptr-Array or Grouped Gemm TMA Warp Specialized Cooperative policies");
};