[mxfp8 moe training] integrate triton quant/dequant kernels into mxfp8 all to all

stack-info: PR: #3197, branch: danielvegamyhre/stack/79

Author: danielvegamyhre

torchao/prototype/moe_training/kernels/mxfp8/comms.py

@@ -11,7 +11,10 @@
     blockwise_barrier,
     sync_threads,
 )
-from torchao.prototype.mx_formats.config import ScaleCalculationMode
+from torchao.prototype.mx_formats.kernels import (
+    triton_mxfp8_dequant_dim0,
+    triton_to_mxfp8_dim0,
+)
 from torchao.prototype.mx_formats.mx_tensor import to_dtype, to_mx
 
 
@@ -473,11 +476,9 @@ def forward(
         """
         # Quantize input
         block_size = 32
-        input_scales, input_data = to_mx(
+        input_data, input_scales = triton_to_mxfp8_dim0(
             input,
-            elem_dtype=torch.float8_e4m3fn,
-            block_size=block_size,
-            scaling_mode=ScaleCalculationMode.RCEIL,
+            inner_block_size=block_size,
         )
 
         # Dispatch data (async)
@@ -501,20 +502,17 @@ def forward(
         output_data = torch.ops._c10d_functional.wait_tensor(output_data)
 
         # Dequantize output
-        lowp_dtype = output_data.dtype
         hp_dtype = input.dtype
-        hp_output = to_dtype(
+        triton_hp_output = triton_mxfp8_dequant_dim0(
             output_data,
-            output_scales.view(torch.float8_e8m0fnu),
-            lowp_dtype,
-            block_size,
+            output_scales,
             hp_dtype,
+            block_size,
         )
-
         ctx.input_splits = input_splits
         ctx.output_splits = output_splits
         ctx.group = group
-        return hp_output
+        return triton_hp_output
 
     @staticmethod
     def backward(ctx, grad_output_hp):
@@ -529,11 +527,9 @@ def backward(ctx, grad_output_hp):
 
         # Quantize grad_output
         block_size = 32
-        grad_out_scales, grad_out_data = to_mx(
+        grad_out_data, grad_out_scales = triton_to_mxfp8_dim0(
             grad_output_hp,
-            elem_dtype=torch.float8_e4m3fn,
-            block_size=block_size,
-            scaling_mode=ScaleCalculationMode.RCEIL,
+            inner_block_size=block_size,
         )
 
         # Dispatch data (async)
@@ -557,13 +553,11 @@ def backward(ctx, grad_output_hp):
         grad_input_scales = torch.ops._c10d_functional.wait_tensor(grad_input_scales)
 
         hp_dtype = grad_output_hp.dtype
-        lowp_dtype = grad_input_data.dtype
-        grad_input_hp = to_dtype(
+        grad_input_hp = triton_mxfp8_dequant_dim0(
             grad_input_data,
-            grad_input_scales.view(torch.float8_e8m0fnu),
-            lowp_dtype,
-            block_size,
+            grad_input_scales,
             hp_dtype,
+            block_size,
         )
         return grad_input_hp, None, None, None