Kim Walisch
Kim Walisch
The current default ```WITH_DIV32=ON``` increases branch mispredictions and only makes sense on CPUs with slow 64-bit integer division.
We currently use libdivide because integer division is very slow on most x64 CPUs. However integer division will likely become much faster in near future and then WITH_LIBDIVIDE=OFF is expected...
I started primesieve in 2010 and primecount in 2013. Since the beginning, my goal for primesieve and primecount was to write libraries that are extremely specialized i.e. primesieve should generate...
```primesieve::iterator```'s performance depends heavily on the ```fillNextPrimes()``` method from ```PrimeGenerator.cpp```. For x64 we have a [vectorized AVX512 algorithm](https://github.com/kimwalisch/primesieve/blob/c2c56e5b01478f6cf8cbfc10443156c482362fab/src/PrimeGenerator.cpp#L349) that is pretty optimal for this task. Once other CPU architectures (e.g....
I tried AVX512 pre-sieving using the 2 algorithms below. On AMD EPYC 4th gen CPUs (Genoa) I saw no speedup using both GCC and Clang (compared to the default SSE2...
Hi Dana, I have seen that you have significantly reworked your ```legendre_phi(x, a)``` implementation on the weekend. This algorithm is also quite important in primecount because it is used in...
> The ```__AVX512F__```, ```__AVX512CD__```, ```__AVX512BW__```, ```__AVX512DQ__``` and ```__AVX512VL__``` preprocessor symbols are defined when the ```/arch:AVX512``` compiler option is specified. According to the Microsoft MSVC website https://learn.microsoft.com/en-us/cpp/build/reference/arch-x64?view=msvc-170 the support of AVX512...
After having tested primesieve with the new Intel Arrow Lake CPUs, here is the new pull request for updating the primesieve test profile to the [latest primesieve-12.6 release](https://github.com/kimwalisch/primesieve/releases). Similar to...
Checklist for Pull Requests - [x] Have you signed the [Contributor License Agreement](https://cla.opensource.microsoft.com/microsoft/winget-pkgs)? Manifests - [x] Have you checked that there aren't other open [pull requests](https://github.com/microsoft/winget-pkgs/pulls) for the same manifest...