Update blog

Author: cmdr2

content/blog/2025-10-10-1760088945.md

@@ -11,6 +11,7 @@ tags:
 Some notes on machine-learning compilers, gathered while researching tech for Easy Diffusion's next engine (i.e. sdkit v3). For context, see the [design constraints](https://cmdr2.github.io/notes/2025/10/1760085894/) of the new engine.
 
 ## tl;dr summary
+
 Caveat: This analysis could change in the future.
 
 The current state is:
@@ -22,6 +23,7 @@ The focus of cross-vendor compilers seems to be either on datacenter hardware, o
 The idea of a cross-vendor ML compiler is clearly awesome, and I think this is the way things should go. But we're not there yet for desktops/laptops, in terms of runtime performance.
 
 ## What's an ML compiler?
+
 The basic idea of an ML compiler is to consider an ML model's execution graph as a program to compile, and produce an optimized set of GPU-specific instructions. The compiler can optimize the execution graph by doing things like fusing operations together, parallelizing operations when possible, and even mapping groups of operators to GPU-specific instructions. It can use its knowledge of the target GPU for optimizing the memory layout and parallelism of operations. Basically what compilers already do for CPUs today, but for GPUs.
 
 We already have a decent graph format: ONNX. Every model that I intend to support has ONNX exports available (and it's easy to export one, for new models).
@@ -34,6 +36,7 @@ We already have a decent graph format: ONNX. Every model that I intend to suppor
 * [https://www.modular.com/blog/democratizing-ai-compute-part-6-what-about-ai-compilers](https://www.modular.com/blog/democratizing-ai-compute-part-6-what-about-ai-compilers)
 
 ## ML compiler projects
+
 Cross-vendor ML compilers:
 - [XLA](https://openxla.org/xla), 2017 (the first major ML compiler)
 - [Apache TVM](https://tvm.apache.org/), 2019