搜索资源列表
vec2d
- 1D/2D 向量计算封装，特别经过SSE优化，能够提供与Intel IPP速度接近的向量计算，支持矩阵转置、矩阵加减乘除。精度支持单/双精度设置，需要SSE2以上指令支持。
sha1-sse2-v2.tar
- SHA1: an older version of the SHA1 code above, and a descr iption of how the technique works.
NASMmanualinchinese
- NASM是一个为可移植性与模块化而设计的一个80x86的汇编器。它支持相当多 的目标文件格式，包括Linux和 NetBSD/FreeBSD , a.out , ELF , COFF ,微软16 位的 OBJ 和 Win32 。它还可以输出纯二进制文件。它的语法设计得相当的简 洁易懂，和Intel语法相似但更简单。它支持 Pentium , P6 , MMX , 3DNow! , SSE and SSE2 指令集-NASM is a modular designed for t
asmjit
- x86指令模擬器，可在C++開發環境下模擬x86/x64指令的運行-AsmJit is a complete x86/x64 JIT Assembler for C++ language. It supports 32-bit and 64-bit mode, FPU, MMX, 3dNow, SSE, SSE2, SSE3 and SSE4 through type-safe API which mimics an Intel assembler syntax and eliminates n
sse2
- SSE-2 implementation of RAID-6 syndrome functions
dct_sse2
- vp8 short fdct 4x4 sse2 for Embedded Linux.
quantize_sse2
- vp8 regular quantize impl sse2 for Linux.
fwalsh_sse2
- vp8 short walsh 4x4 sse2 for Embedded Linux.
variance_impl_mmx
- vp8 get mb ss sse2 for Embedded Linux.
subtract_sse2
- vp8 subtract sse2 impl for Linux.
recon_sse2
- vp8 recon4b sse2 for Linux.
iwalsh_sse2
- vp8 short inv walsh 4x4 sse2 for Embedded Linux.
subpixel_sse2
- vp8 filter block 1d8_h6 sse2 for Embedded Linux.
serpent-sse2-x86_64-asm_64
- Serpent Cipher 8-way parallel algorithm (x86_64 SSE2).
serpent_sse2_glue
- Glue Code for SSE2 assembler versions of Serpent Cipher.
serpent-sse2-i586-asm_32
- Serpent Cipher 4-way parallel algorithm (i586 SSE2).
serpent-sse2
- Common hwbus abstraction layer interface for cw1200 wireless driver.
sse2
- SSE-2 implementation of RAID-6 syndrome.
SkBitmapProcState_opts_SSE2
- SSE2 only support 16bit interger max & min, so only process the case maxX less than the max 16bit interger. -SSE2 only support 16bit interger max & min, so only process the case maxX less than the max 16bit interger.
vp9_variance_impl_intrin_avx2
- processing two strides in a 256 bit register reducing the number of loop stride by half (comparing to the sse2 code). -processing two strides in a 256 bit register reducing the number of loop stride by half (comparing to the sse2 code).