[BE/docs] Add fp8 rowwise perf table to float8 training readme (#2312)

* add fp8 rowwise perf table

* add fp8 rowwise perf table to readme

Author: danielvegamyhre

docs/static/fp8-rowwise-perf.png

@@ -132,7 +132,9 @@ on using `torchao.float8` in a distributed setting.
 
 # Performance
 
-A common question about float8 training is "when is float8 linear faster vs bfloat16?".  Given the M, K, N of the forward pass through your linear, you can reference the table below for a microbenchmark based speedup estimate on NVIDIA H100:
+A common question about float8 training is "when is float8 linear faster vs bfloat16?".  Given the M, K, N of the forward pass through your linear, you can reference the tables below for a microbenchmark based speedup estimate on NVIDIA H100:
+
+### Tensorwise scaling
 
 <img width="805" alt="float8_speedup" src="https://github.com/user-attachments/assets/5c5f2817-7eb7-4cab-bd03-49fe70cd31a8">
 
@@ -152,6 +154,11 @@ To reproduce the raw data for table above, you can run the following script
 python benchmarks/float8/float8_roofline.py your_output_filename.csv --shape_gen_name sweep
 ```
 
+### Rowwise scaling
+
+<img width="805" alt="float8_rowwise_speedup" src="../../docs/static/fp8-rowwise-perf.png" />
+
+
 ## Derivation
 
 In a bf16 linear, assume all of the time is spent in gemms.  In a float8 linear, account for max_abs and casting overhead.  We want to know when