Files
leonarski_fandClaude Opus 5.5 3d88ebea65 Refinement on the GPU: the normal matrix's budget counts J's copy in the parameter atoms' order
Since the transpose product reads J from a per-atom ordered copy (jt) and each entry's row from
row_at, an entry holds two float4 and three int (param, entry, row_at) on the card; row_of stays on
the host.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UqXYdHHZ4cogS1mrGUtfN7
2026-11-17 16:59:38 +01:00

99 lines
4.7 KiB
C++

// SPDX-FileCopyrightText: 2026 Filip Leonarski, Paul Scherrer Institute <filip.leonarski@psi.ch>
// SPDX-License-Identifier: GPL-3.0-only
#include "RefinementNewtonGPU.h"
#include "RefinementRestraints.h"
size_t RestraintNormalMatrixGPU::DeviceBytes(const RestraintSet &set, size_t n_atoms,
const std::vector<int> &block_atoms, int iterations) {
size_t entries = 0, rows = 0, params = 0, inverse = 0;
for (const RestraintRecord &r : set.records) // as At() lays out J
switch (static_cast<RestraintKind>(r.kind)) {
case RestraintKind::Plane:
entries += static_cast<size_t>(r.atom[1]) * r.atom[1];
rows += r.atom[1];
break;
case RestraintKind::Angle:
entries += 3;
rows++;
break;
case RestraintKind::Torsion:
case RestraintKind::Chiral:
case RestraintKind::NcsDistance:
entries += 4;
rows++;
break;
default:
entries += 2;
rows++;
break;
}
for (int a : block_atoms) {
params += a;
inverse += static_cast<size_t>(3 * a) * (3 * a + 1) / 2;
}
const size_t n = 3 * params;
return n_atoms * 4 * sizeof(float) // the positions
+ set.records.size() * (sizeof(RestraintRecord) + sizeof(int)) + set.plane_atoms.size() * sizeof(int)
+ rows * (sizeof(int) + 2 * sizeof(float)) // row starts, J v
+ entries * (8 * sizeof(float) + 3 * sizeof(int)) + params * sizeof(int) // J, its transposed copy, index
+ n * 15 * sizeof(float) // seven vectors, diagonal
+ n * (2 * sizeof(float) + 2 * sizeof(int)) + 2 * block_atoms.size() * sizeof(int) // h, the blocks
+ inverse * 2 * sizeof(float) // their inverses
+ static_cast<size_t>(iterations) * n * 3 * sizeof(float); // the kept residuals
}
// The host half: the index of J, as RestraintGaussNewton lays out its rows, with every row kept.
void RestraintNormalMatrixGPU::At(const RestraintSet &set, const std::vector<std::array<double, 4>> &xyzb) {
if (!indexed_) {
std::vector<int> record_entry, start{0}, atom;
for (const RestraintRecord &r : set.records) {
record_entry.push_back(static_cast<int>(atom.size()));
switch (static_cast<RestraintKind>(r.kind)) {
case RestraintKind::Plane:
// one row per atom of the plane, each over all of them
for (int i = 0; i < r.atom[1]; i++) {
for (int k = 0; k < r.atom[1]; k++)
atom.push_back(set.plane_atoms[r.atom[0] + k]);
start.push_back(static_cast<int>(atom.size()));
}
continue;
case RestraintKind::Angle:
atom.insert(atom.end(), r.atom, r.atom + 3);
break;
case RestraintKind::Torsion:
case RestraintKind::Chiral:
case RestraintKind::NcsDistance:
atom.insert(atom.end(), r.atom, r.atom + 4);
break;
default: // Bond, Repulsion, BSimilarity
atom.insert(atom.end(), r.atom, r.atom + 2);
break;
}
start.push_back(static_cast<int>(atom.size()));
}
const size_t rows = start.size() - 1;
std::vector<int> row_of(atom.size()), param(atom.size());
for (size_t e = 0; e < atom.size(); e++)
param[e] = param_of_[atom[e]];
for (size_t r = 0; r < rows; r++)
for (int e = start[r]; e < start[r + 1]; e++)
row_of[e] = static_cast<int>(r);
std::vector<int> entry_start(n_params_ + 1, 0);
for (int k : param)
if (k >= 0)
entry_start[k + 1]++;
for (size_t k = 0; k < n_params_; k++)
entry_start[k + 1] += entry_start[k];
std::vector<int> entry(entry_start.back());
std::vector<int> fill(entry_start.begin(), entry_start.end() - 1);
for (size_t e = 0; e < param.size(); e++)
if (param[e] >= 0)
entry[fill[param[e]]++] = static_cast<int>(e);
Upload(set.records, set.plane_atoms, record_entry, start, param, row_of, entry_start, entry);
indexed_ = true;
}
Fill(xyzb);
}