Documentation update for FP8 on MI300 (#766)

docs/conceptual/pipeline-metrics.rst

@@ -519,6 +519,13 @@ MFMA instructions are classified by the type of input data they operate on, and
 
      - Instructions per :ref:`normalization unit <normalization-units>`
 
+   * - MFMA-F8 Instructions
+
+     - The total number of 8-bit floating point :ref:`MFMA <desc-mfma>`
+       instructions issued per :ref:`normalization unit <normalization-units>`. This is supported in AMD Instinct MI300 series and later only.
+
+     - Instructions per :ref:`normalization unit <normalization-units>`
+
    * - MFMA-F16 Instructions
 
      - The total number of 16-bit floating point :ref:`MFMA <desc-mfma>`

docs/conceptual/system-speed-of-light.rst

@@ -49,6 +49,16 @@ of ROCm Compute Profiler’s profiling report.
 
      - GIOPs
 
+   * - :ref:`MFMA <desc-mfma>` FLOPs (F8)
+
+     - The total number of 8-bit floating point :ref:`MFMA <desc-mfma>`
+       operations executed per second. This does not include any 16-bit
+       brain floating point operations from :ref:`VALU <desc-valu>`
+       instructions. This is also presented as a percent of the peak theoretical
+       F8 MFMA operations achievable on the specific accelerator. It is supported on AMD Instinct MI300 series and later only.
+
+     - GFLOPs
+
    * - :ref:`MFMA <desc-mfma>` FLOPs (BF16)
 
      - The total number of 16-bit brain floating point :ref:`MFMA <desc-mfma>`