Merge pull request #2 from DanielRichtmann/feature/fused-innerproduct-norm2

Fused innerProduct + norm2 on first argument operation

Grid/lattice/Lattice_reduction.h

@@ -204,8 +204,64 @@ axpby_norm_fast(Lattice<vobj> &z,sobj a,sobj b,const Lattice<vobj> &x,const Latt
   grid->GlobalSum(nrm);
   return nrm; 
 }
-
  
+template<class vobj> strong_inline void
+innerProduct_norm(ComplexD& ip, RealD &nrm, const Lattice<vobj> &left,const Lattice<vobj> &right)
+{
+  conformable(left,right);
+
+  typedef typename vobj::scalar_type scalar_type;
+  typedef typename vobj::vector_typeD vector_type;
+  Vector<ComplexD> tmp(2);
+
+  GridBase *grid = left.Grid();
+
+  auto left_v=left.View();
+  auto right_v=right.View();
+
+  const uint64_t nsimd = grid->Nsimd();
+  const uint64_t sites = grid->oSites();
+
+#ifdef GRID_NVCC
+  // GPU
+  typedef decltype(innerProduct(left_v[0],right_v[0])) inner_t;
+  typedef decltype(innerProduct(left_v[0],left_v[0])) norm_t;
+  Vector<inner_t> inner_tmp(sites);
+  Vector<norm_t> norm_tmp(sites);
+  auto inner_tmp_v = &inner_tmp[0];
+  auto norm_tmp_v = &norm_tmp[0];
+
+  accelerator_for( ss, sites, nsimd,{
+      auto left_tmp = left_v(ss);
+      coalescedWrite(inner_tmp_v[ss],innerProduct(left_tmp,right_v(ss)));
+      coalescedWrite(norm_tmp_v[ss],innerProduct(left_tmp,left_tmp)));
+  });
+
+  tmp[0] = TensorRemove(sumD_gpu(inner_tmp_v,sites));
+  tmp[1] = TensorRemove(sumD_gpu(norm_tmp_v,sites));
+#else
+  // CPU
+  typedef decltype(innerProductD(left_v[0],right_v[0])) inner_t;
+  typedef decltype(innerProductD(left_v[0],left_v[0])) norm_t;
+  Vector<inner_t> inner_tmp(sites);
+  Vector<norm_t> norm_tmp(sites);
+  auto inner_tmp_v = &inner_tmp[0];
+  auto norm_tmp_v = &norm_tmp[0];
+
+  accelerator_for( ss, sites, nsimd,{
+      auto left_tmp = left_v(ss);
+      inner_tmp_v[ss] = innerProductD(left_tmp,right_v(ss));
+      norm_tmp_v[ss] = innerProductD(left_tmp,left_tmp);
+  });
+  // Already promoted to double
+  tmp[0] = TensorRemove(sum(inner_tmp_v,sites));
+  tmp[1] = TensorRemove(sum(norm_tmp_v,sites));
+#endif
+  grid->GlobalSumVector(&tmp[0],2); // keep norm Complex -> can use GlobalSumVector
+  ip = tmp[0];
+  nrm = real(tmp[1]);
+}
+
 template<class Op,class T1>
 inline auto sum(const LatticeUnaryExpression<Op,T1> & expr)
   ->typename decltype(expr.op.func(eval(0,expr.arg1)))::scalar_object

tests/Test_innerproduct_norm.cc

@@ -0,0 +1,126 @@
+/*************************************************************************************
+
+Grid physics library, www.github.com/paboyle/Grid
+
+Source file: ./tests/Test_innerproduct_norm.cc
+
+Copyright (C) 2015
+
+Author: Daniel Richtmann <daniel.richtmann@ur.de>
+
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 2 of the License, or
+(at your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+GNU General Public License for more details.
+
+You should have received a copy of the GNU General Public License along
+with this program; if not, write to the Free Software Foundation, Inc.,
+51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+
+See the full license in the file "LICENSE" in the top level distribution directory
+*************************************************************************************/
+/*  END LEGAL */
+#include <Grid/Grid.h>
+
+using namespace Grid;
+
+int main(int argc, char** argv) {
+  Grid_init(&argc, &argv);
+
+  const int nIter = 100;
+
+  // clang-format off
+  GridCartesian *Grid_d = SpaceTimeGrid::makeFourDimGrid(GridDefaultLatt(), GridDefaultSimd(Nd, vComplexD::Nsimd()), GridDefaultMpi());
+  GridCartesian *Grid_f = SpaceTimeGrid::makeFourDimGrid(GridDefaultLatt(), GridDefaultSimd(Nd, vComplexF::Nsimd()), GridDefaultMpi());
+  // clang-format on
+
+  GridParallelRNG pRNG_d(Grid_d);
+  GridParallelRNG pRNG_f(Grid_f);
+
+  std::vector<int> seeds_d({1, 2, 3, 4});
+  std::vector<int> seeds_f({5, 6, 7, 8});
+
+  pRNG_d.SeedFixedIntegers(seeds_d);
+  pRNG_f.SeedFixedIntegers(seeds_f);
+
+  // clang-format off
+  LatticeFermionD x_d(Grid_d); random(pRNG_d, x_d);
+  LatticeFermionD y_d(Grid_d); random(pRNG_d, y_d);
+  LatticeFermionF x_f(Grid_f); random(pRNG_f, x_f);
+  LatticeFermionF y_f(Grid_f); random(pRNG_f, y_f);
+  // clang-format on
+
+  GridStopWatch sw_ref;
+  GridStopWatch sw_res;
+
+  { // double precision
+    ComplexD ip_d_ref, ip_d_res, diff_ip_d;
+    RealD    norm2_d_ref, norm2_d_res, diff_norm2_d;
+
+    sw_ref.Reset();
+    sw_ref.Start();
+    for(int i = 0; i < nIter; ++i) {
+      ip_d_ref    = innerProduct(x_d, y_d);
+      norm2_d_ref = norm2(x_d);
+    }
+    sw_ref.Stop();
+
+    sw_res.Reset();
+    sw_res.Start();
+    for(int i = 0; i < nIter; ++i) { innerProduct_norm(ip_d_res, norm2_d_res, x_d, y_d); }
+    sw_res.Stop();
+
+    diff_ip_d    = ip_d_ref - ip_d_res;
+    diff_norm2_d = norm2_d_ref - norm2_d_res;
+
+    // clang-format off
+    std::cout << GridLogMessage << "Double: ip_ref = " << ip_d_ref << " ip_res = " << ip_d_res << " diff = " << diff_ip_d << std::endl;
+    std::cout << GridLogMessage << "Double: norm2_ref = " << norm2_d_ref << " norm2_res = " << norm2_d_res << " diff = " << diff_norm2_d << std::endl;
+    std::cout << GridLogMessage << "Double: time_ref = " << sw_ref.Elapsed() << " time_res = " << sw_res.Elapsed() << std::endl;
+    // clang-format on
+
+    assert(diff_ip_d == 0.);
+    assert(diff_norm2_d == 0.);
+
+    std::cout << GridLogMessage << "Double: all checks passed" << std::endl;
+  }
+
+  { // single precision
+    ComplexD ip_f_ref, ip_f_res, diff_ip_f;
+    RealD    norm2_f_ref, norm2_f_res, diff_norm2_f;
+
+    sw_ref.Reset();
+    sw_ref.Start();
+    for(int i = 0; i < nIter; ++i) {
+      ip_f_ref    = innerProduct(x_f, y_f);
+      norm2_f_ref = norm2(x_f);
+    }
+    sw_ref.Stop();
+
+    sw_res.Reset();
+    sw_res.Start();
+    for(int i = 0; i < nIter; ++i) { innerProduct_norm(ip_f_res, norm2_f_res, x_f, y_f); }
+    sw_res.Stop();
+
+    diff_ip_f    = ip_f_ref - ip_f_res;
+    diff_norm2_f = norm2_f_ref - norm2_f_res;
+
+    // clang-format off
+    std::cout << GridLogMessage << "Single: ip_ref = " << ip_f_ref << " ip_res = " << ip_f_res << " diff = " << diff_ip_f << std::endl;
+    std::cout << GridLogMessage << "Single: norm2_ref = " << norm2_f_ref << " norm2_res = " << norm2_f_res << " diff = " << diff_norm2_f << std::endl;
+    std::cout << GridLogMessage << "Single: time_ref = " << sw_ref.Elapsed() << " time_res = " << sw_res.Elapsed() << std::endl;
+    // clang-format on
+
+    assert(diff_ip_f == 0.);
+    assert(diff_norm2_f == 0.);
+
+    std::cout << GridLogMessage << "Single: all checks passed" << std::endl;
+  }
+
+  Grid_finalize();
+}