ErmIg questions

ErmIg answers

• How to force use of vmovapd instead of vmovupd when programming SIMDed code using intel intrinsics?
• Convert 8 16 bit SSE register to 8bit data
• Detecting how blurred an image is
• AVX-512BW emulation of _mm512_dpbusd_epi32 AVX-512VNNI instruction
• Vectorization of modulo multiplication
• Fast implementation of covariance of two 8-bit arrays
• Automatic vectorization with g++ of a loop with bit operations
• How to add an AVX2 vector horizontally 3 by 3?
• set individual bit in AVX register (__m256i), need "random access" operator
• Dot Product of Vectors with SIMD