Parallelism_level

[reymono]

Nice to see Mojo reach v.1.0

I’m comparing Mojo to C on a mandelbrot task.

This on a Macbook Air M5. SIMD works well, but multi-threaded performance suffers and I think it is because parallelism_level()is limited to 4. Is there a way to increase it?

The M5 has 4 performance cores and 6 efficiency cores - performance cores are faster. However, the C program is significantly faster when the number of threads is bumped up to 9 or 10. I assume the same would be the case for Mojo.

Language	Repository	Single Thread	Multi-Thread	Simd	Multi-Thread + Simd
C	mandelbrot-c	3.6	0.6
Mojo	mandelbrot-mojo	3.8	1.2	0.7	0.4

[AlisterBurt]

Hi @reymono ! We’re excited about 1.0 too

The short answer is no, this can’t be configured right now. Here’s a message from @owenhilyard explaining the current state of threading and some of the reasons it hasn’t been added yet.

How to multithreading & synchronize state (CPU)

The current state of async in Mojo is that it would be generous to call it “half baked”. We ran into a bunch of language level issues and Mojo does not currently have a formal threading model, primarily because we need to determine a threading model that works almost universally, applying not just to CPUs but also to GPUs and exotic ASICs. There are other things which are higher priority to fix before 1.0, such as traits, closures and the type system, so async is likely going to take a while. N…