syoyo

// A64FX SVE 1.0 High-Performance FP16 GEMM with FP32 Accumulation
// Target: ~160% of FP32 peak (~80% of FP16 peak)
//
// Strategy:
// - Use FP16 FMLA directly for maximum throughput (32 ops/instruction)
// - Accumulate in FP16 for K_INTERVAL iterations
// - Convert to FP32 and add to FP32 accumulators periodically
// - Software pipeline: overlap loads, converts, and FMAs
//
// A64FX FP16 peak: ~264 GFLOPS/core (2x FP32)

syoyo

#include <iostream>
  #include <string>

  static int _ReadUTF8(char const *&cp, std::string *errMsg)
  {
      // Return a byte with the high `n` bits set, rest clear.
      auto highBits = [](int n) {
          return static_cast<unsigned char>(((1 << n) - 1) << (8 - n));
      };

syoyo

#!/bin/bash

VERSION="1.0.2"
# Interface connect to out lan
INTERFACE="eth0"
# Interface virtual for incomming traffic
VIRTUAL="ifb0"
# set the direction (1 = outgoing only, 2 = incoming only 3 = both)
DIRECTION=3
# Speed

syoyo

// Issue
// * Allocate and provice insufficient data size(e.g. vertex data) fo GAS input.
// * Call `optixAccelBuild` with "correct" buffer size(e.g. `buildInput.triangleArray.numVertices`)
//    * Build GAS against invalid data but no error report at this time
// * Some memory object causes 'illegal memory access' error when freeing(also happens in CUDA driver API)

// How to reprocible
//
// OptiX 7.1 optixHair SDK sample
// Uncomment following in Util.h

syoyo

#include <arm_neon.h>
#include <cstdio>
#include <limits>
#include <cassert>
#include <cmath>
#include <cstdint>

bool check_snan(float f)
{
  bool is_nan = std::isnan(f);

syoyo

#include <arm_neon.h>
#include <cstdio>
#include <limits>
#include <cassert>
#include <cmath>
#include <cstdint>

bool check_snan(float f)
{
  bool is_nan = std::isnan(f);

syoyo

//
// C++ implementaion of "A simple method to construct isotropic quasirandom blue
// noise point sequences"
//
// http://extremelearning.com.au/a-simple-method-to-construct-isotropic-quasirandom-blue-noise-point-sequences/
//

// Assume 0 <= x
static double myfmod(double x) { return x - std::floor(x); }

syoyo

// Writing (&x)[i] in operator[] is safe/correct C++ code or not? 
// The following is the reduced code fragment(Thus not work by just copy&paste) which Intel C++ compier(ver 13 and 15) miscompiles(Release build only) the code for the access to real3 object through operator[] inside OpenMP loop.
// clang and gcc are OK to compile&run

typedef float real;

struct real3 {
  real3() {}
  real3(real xx, real yy, real zz) {
    x = xx;

syoyo

#!/bin/bash

rm -rf CMake*

export NDK=/home/syoyo/local/android-ndk-r10e
export SYSROOT=$NDK/platforms/android-21/arch-arm64

export CC="$NDK/toolchains/aarch64-linux-android-4.9/prebuilt/linux-x86_64/bin/aarch64-linux-android-gcc"

export CXX="$NDK/toolchains/aarch64-linux-android-4.9/prebuilt/linux-x86_64/bin/aarch64-linux-android-g++"

syoyo

#!/usr/bin/env python

import ninja_syntax
import glob
import os

cxx_files = [
    "src/OptionParser.cpp"
  , "src/easywsclient.cpp"
  , "src/json_to_eson.cc"