CUTLASS 3.3.0 (#1167)

* Release 3.3.0

Adds support for mixed precision GEMMs On Hopper and Ampere
Adds support for < 16B aligned GEMMs on Hopper
Enhancements to EVT
Enhancements to Python interface
Enhancements to Sub-byte type handling in CuTe
Several other bug-fixes and performance improvements.

* minor doc update

tools/profiler/src/device_allocation.cu

@@ -2005,7 +2005,6 @@ void DeviceAllocation::write_tensor_csv(
   case library::NumericTypeID::kFE5M2:
     write_tensor_csv_static_type<float_e5m2_t>(out, *this);
     break;
-
   case library::NumericTypeID::kF16:
     write_tensor_csv_static_type<half_t>(out, *this);
     break;

tools/profiler/src/gemm_operation_profiler.cu

@@ -585,11 +585,11 @@ Status GemmOperationProfiler::initialize_workspace(
       workspace_size = underlying_operation->get_device_workspace_size(&gemm_workspace_.configuration,
                                                             &gemm_workspace_.arguments);
       gemm_workspace_.device_workspace.reset(library::NumericTypeID::kU8, workspace_size);
+
       status = underlying_operation->initialize(
         &gemm_workspace_.configuration,
         gemm_workspace_.host_workspace.data(),
         gemm_workspace_.device_workspace.data());
-
       if (status != Status::kSuccess) {
         return status;
       }