Hi, I'm new here, so I'm not sure how to expose my question.

Currently, I'm working on a mode 7 engine for casio FX9860 series. However, the code my co-programmer and I have come up with doesn't run at a satisfying speed, even after hours of optimization.

So I'm looking for a way to implement assembly code in a Casio SDK project.
Does anyone here know how to achieve this?

If I remember correctly the SDK uses GCC so this should show you how to use assembly http://www.codeproject.com/Articles/15971/Using-Inline-Assembly-in-C-C

The official Casio SDK for the fx9860G doesn't use GCC, but Hitachi's C compiler.

A skim of the Renesas docs didn't reveal any support for inline assembly, but you can still link code written in assembly against code written in C. Most of the immediately relevant information appears to be in section 9.3.

That said, I don't recommend micro-optimizing like this unless you have evidence that you'll be able to speed it up enough. Examine the code generated by the compiler in your hot sections, and only rewrite it if it's definitely subpar. Otherwise you're just wasting effort that would be better spent using better algorithms.