add dim order into executorch concept documentation

Summary: as title

Differential Revision: D60081081

docs/source/concepts.md

@@ -93,6 +93,14 @@ An interface that enables the ExecuTorch runtime to read from a file or other da
 
 To run parts (or all) of a program on a specific backend (eg. XNNPACK) while the rest of the program (if any) runs on the basic ExecuTorch runtime. Delegation enables us to leverage the performance and efficiency benefits of specialized backends and hardware.
 
+## Dim Order
+
+Dimension order, the ExecuTorch way to describe tensor's memory format by explicitly stating the order of dimensions from outermost to innermost.
+
+For example, for the [N, H, W, C], or [channels_last memory format](https://pytorch.org/tutorials/intermediate/memory_format_tutorial.html), we use [0, 2, 3, 1] to record its dim order.
+
+Currently ExecuTorch only supports dim order representation for [contiguous](https://pytorch.org/blog/tensor-memory-format-matters/) and [channels_last](https://pytorch.org/tutorials/intermediate/memory_format_tutorial.html) memory format.
+
 ## DSP (Digital Signal Processor)
 
 Specialized microprocessor chip with architecture optimized for digital signal processing.