manupa-arm commented on a change in pull request #65:
URL: https://github.com/apache/tvm-rfcs/pull/65#discussion_r839363818



##########
File path: rfcs/0009_Unified_Static_Memory_Planning.md
##########
@@ -515,4 +663,6 @@ NOTE : to support tir.constants generally, we'll be 
enhancing the bound relay.co
 
 # Drawbacks
 
-* The relay "main" function that describes the call order to operator 
PrimFuncs has to be described in TIR to be able to integrate the USMP into the 
respective executor codegen. However, we dont view this as a major problem as 
the relay "main" function could easily be lowered to TIR.
\ No newline at end of file
+* The relay "main" function that describes the call order to operator 
PrimFuncs has to be described in TIR to be able to integrate the USMP into the 
respective executor codegen. However, we dont view this as a major problem as 
the relay "main" function could easily be lowered to TIR.
+
+* The U4 usecase will only be supported with [Embedded C Runtime 
Interface](https://discuss.tvm.apache.org/t/rfc-utvm-embedded-c-runtime-interface/9951/14).
 This is mainly because the nature of the requirement is associated with 
embedded usecases. However, the USMP changes here should be complimentary to 
support other runtime interfaces such as Module-based Model Runtime Interface's 
set_input and set_output in future.

Review comment:
       Good point again!
   
   This is generally applicable for Embedded C Runtime Inferface as well.
   
   Another way to frame the suggestion is to say : why cant we treat the buffer 
space used by I/O tensors as seperate workspace buffers ? 
   
   I have considered this too :).  The short anwser is it is sub-optimal use 
disjoint workspace buffers rather than a single workspace buffer -- because it 
does not allow partial overlaps of tensors across such buffers. TLDR; 
   
   So the workspace buffer is viewed as a spatio-temporal buffer and tensors 
are considered as spatio-temporal allocation. Having disjoint workspace 
buffers, create a spatial barrier that could potentially lead to higher 
fragmentation. Therefore, out of the two options; we are prioritizing the 
better one here. Again, this could be an another usecase if there is value for 
it.
   
   Again, I can add few lines for this as future consideration.
   
   




-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


Reply via email to