Programming-Language-Benchmark VS weave

I found Zig implementation of json parsing is interesting. The code is free from hidden control flow !.

https://github.com/hanabi1224/Programming-Language-Benchmark...

It appears helloworld is the only test with any repeats, and it only has 5 repeats. https://github.com/hanabi1224/Programming-Language-Benchmark...

Here's the measurement code, it appears to be significantly more complicated than a simple fork/exec/wait loop but that could just be all the C# getting in the way: https://github.com/hanabi1224/Programming-Language-Benchmark... Nevertheless you are probably right that the bulk of this 1.8ms is in the executable under test, and it truly is just bloat. Running `hyperfine ./empty-main-function` from rustc on my Mac gives 0.8ms.

Is faster... on code that has been optimized to hell and back 5 times over and no longer resembles anything like normal code written in the language.

Seriously, this is the code for the top program. I'm reasonably sure 99% of C++ programmers could not decipher it without spending significant amounts of time on google: https://github.com/hanabi1224/Programming-Language-Benchmark...

I appreciate that fair benchmarks across languages are a hard problem, but this is not a good solution to it. Any reference to this data as a comparison between "programming languages and compilers" needs to come with a giant disclaimer that it's comparing them at something you almost certainly don't use them for, and is very far from their main use case.

I also appreciate that this is a repetitive comment the likes of which always come up when this benchmark is mentioned... but I really don't see another way to avoid people misinterpreting it. Very few people are going to spontaneously click through to the code.

GPU drivers provide an event system:

- Cuda: https://github.com/mratsim/weave/issues/133

```

Note: the Theoretical peak limit is hardcoded and used my previous machine i9-9980XE.

It maybe that your BLAS library is not named libopenblas.so, you can change that here: https://github.com/mratsim/laser/blob/master/benchmarks/thir...

Implementation is in this folder: https://github.com/mratsim/laser/tree/master/laser/primitive...

in particular, tiling, cache and register optimization: https://github.com/mratsim/laser/blob/master/laser/primitive...

AVX512 code generator: https://github.com/mratsim/laser/blob/master/laser/primitive...

And generic Scalar/SSE/AVX/AVX2/AVX512 microkernel generator (this is Nim macros to generate code at compile-time): https://github.com/mratsim/laser/blob/master/laser/primitive...

I'll come back later with details on how to use my custom HPC threadpool Weave instead of OpenMP (https://github.com/mratsim/weave/tree/master/benchmarks/matm...)

In my benchmarks, Nim is faster than Rust:

- multithreading runtime (i.e Rayon vs Weave https://github.com/mratsim/weave)

- Cryptography: https://hackmd.io/@gnark/eccbench#Pairing

- Scientific computing / matrix multiplication: https://github.com/bluss/matrixmultiply/issues/34#issuecomme...

There is no inherent reason why a Nim program would be slower than Rust.

If you're interested into diving into this I have reviewed solutions to cactus stacks / split stacks here https://github.com/mratsim/weave/blob/master/weave/memory/multithreaded_memory_management.md

[4] https://github.com/mratsim/weave