simdeez
simdeez copied to clipboard
arduano
easy simd
While AVX2's `_mm_fmsub_*` intrinsics subtract the last operand from the product, ARM's `vmsfq_*` intrinsics subtract the product from the last operand. This means we can use them to implement a...
The Cargo reference [recommends that features be additive](https://doc.rust-lang.org/cargo/reference/features.html#feature-unification), which means that enabling a feature doesn't disable any functionality. As such, they recommend an opt-out `std` feature instead of an opt-in...
Fixes the `Code Checks (formatting, clippy)` GitHub action.
Adds support for WebAssembly SIMD, using the currently available stable intrinsics in Rust. I also added a GitHub action for automated testing. Benchmarking also works, and shows a ~3x improvement...
@arduano did an overhaul of the SIMD traits. That was a big undertaking and has many nice improvements like operator overloading. I am trying to port my stuff over to...
I started working on transitioning (https://github.com/verpeteren/rust-simd-noise/pull/50/files#top) from `rust-simd-noise` on top of the 64 bit changes (https://github.com/verpeteren/rust-simd-noise/pull/51), but got kindof stuck because it seems that the `Avx2`, `Sse2`, `Sse41` engines are...
Hi! Is there a particular reason for which only 16- and 32-bits wide integer vectors are implemented, but not the 8-bits wide ones?
Does simdeez support Apple Silicon? The M1/M2 macs do have support for the AMD SIMD instruction set, just not `avx`.
As discussed in [this issue](https://github.com/arduano/simdeez/issues/34) it would be better that the `simd_xxx_generate` macros are explictly `unsafe`. This has an impact for consumers of this crate, so a (minor) version bump...
Explicit generic parameters aren't parsed, and `impl Trait` arguments don't work due to the generic parameter the macro adds itself. This makes it difficult to abstract SIMD code. I can't...
