Due to some limitation, I can't use CefSharp in my main application, instead I start a separate hosting process to do offscreen rendering and use .NET remoting and memory mapped file to copy the buffer.

The whole procedure is as follows:

1. Host process do offscreen rendering
2. copy rendered buffer to shared memory
3. fire an event with .net remoting
4. client render the resulting bitmap from shared memory (using direct2d api)

The performance was acceptable, but I wonder if this can be done more efficiently? Namely, directly render to my shared memory region in host process, saving additional copy overhead in step 2.

CodePudding user response：

Namely, directly render to my shared memory region in host process, saving additional copy overhead in step 2.

No, that's not possible.

Performance wise only copy the dirty rectangle.