llvm

llvm copied to clipboard

Signed-off-by: Mariusz Merecki [email protected]

@bashbaug @AlexeySotkin @kvladimi @aus-intel could you please take a look at this MR?

@AlexeySachkov, @bader - when we can expect this extension to be reviewed/merged? Some of ESIMD features depend on this.

Have you given any thought how similar float controls could apply to bfloat16 or tf32 types, which would have the same "target width" as float and half types respectively?

yes, I thought we could replace "target width" with an enum whose values would be float = 32, half = 16, double = 64 and something new for bfloat16 or tf32. Such change would be backwards compatible.

yes, I thought we could replace "target width" with an enum whose values would be float = 32, half = 16, double = 64 and something new for bfloat16 or tf32. Such change would be backwards compatible.

Nice, this seems reasonable.

Since the "target width" is an unsigned 32-bit integer maybe it makes sense to choose absurdly large values (e.g. with the high bit set) to represent alternate floating-point encodings like bfloat16 or tf32?

This pull request is stale because it has been open 180 days with no activity. Remove stale label or comment or this will be automatically closed in 30 days.

This pull request was closed because it has been stalled for 30 days with no activity.

@mmerecki @bashbaug hi, do you know, what is the status here?