CUTLASS 3.6.0 (#1850)

* v3.6

* update changelog

* update readme

* fix typo

* fixing typos

* hopper gemm with weight prefetch

---------

Co-authored-by: yuzhai <yuzhai@nvidia.com>
Co-authored-by: Haicheng Wu <haichengw@nvidia.com>

include/cute/arch/util.hpp

@@ -31,7 +31,6 @@
 #pragma once
 
 #include <cute/config.hpp>
-
 #include <cute/numeric/integer_sequence.hpp>
 
 #if defined(__clang__) && defined(__CUDA__)
@@ -254,6 +253,28 @@ explode(Fn fn,
   return fn(d[Id]..., a[Ia]..., b[Ib]..., c[Ic]..., e[Ie]..., f[If]...);
 }
 
+template <class Fn,
+          class PtrD, int... Id,
+          class PtrA, int... Ia,
+          class PtrB, int... Ib,
+          class PtrC, int... Ic,
+          class PtrE, int... Ie,
+          class PtrF, int... If,
+          class PtrG, int... Ig>
+CUTE_HOST_DEVICE constexpr
+void
+explode(Fn fn,
+        PtrD&& d, int_sequence<Id...>,
+        PtrA&& a, int_sequence<Ia...>,
+        PtrB&& b, int_sequence<Ib...>,
+        PtrC&& c, int_sequence<Ic...>,
+        PtrE&& e, int_sequence<Ie...>,
+        PtrF&& f, int_sequence<If...>,
+        PtrG&& g, int_sequence<Ig...>)
+{
+  return fn(d[Id]..., a[Ia]..., b[Ib]..., c[Ic]..., e[Ie]..., f[If]..., g[Ig]...);
+}
+
 //
 // Utility for exploding tuples into functions
 //