leonarski_f
ea383f9e83
A port of the fast-feedback indexer's CUDA kernels to plain C++/Eigen: the same half-sphere and rotation samples, the same scores and the same 32 output cells for Refine(). The loops are reorganised so the inner one runs over samples and vectorises (GCC/Clang, no intrinsics), and the cell search separates the rotation about the candidate vector out of the per-spot terms. The vector refinement reproduces the library's warp sums - partial sums its shuffle reduction leaves in the other lanes carry into the next round - and its float arithmetic (FMA contraction, the Fibonacci points in float), because the candidates searched depend on it: on the library's 27 test spot lists and 4 synthetic stills the CPU and GPU best cells agree to ~1e-6 relative, the top-32 sets in 30 of 31 cases completely. FFBIDX now resolves to the GPU library where a GPU is visible and to the CPU port otherwise (no GPU, --no-gpu, or a build without CUDA); Auto takes FFBIDX for stills whenever a cell is known, also without a GPU, instead of FFTW. With a GPU nothing changes (a stills set and a rotation set reproduce the baseline exactly). Without a GPU the stills indexer pool gets one indexer per worker instead of eight, since the CPU indexer holds no card memory. Single core, synthetic still, loaded 5950X: ~8 ms per call at 30 spots, ~15 ms at 80, ~35 ms at 250; 16 threads ~1300/700/330 calls/s. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015CeEGgpKdfkCTJb3YNMJcu (cherry picked from commit 9a8eb8669d1be6c3ec16db6786280affd6830ee5)
31 lines
1.1 KiB
C++
31 lines
1.1 KiB
C++
// SPDX-FileCopyrightText: 2026 Filip Leonarski, Paul Scherrer Institute <filip.leonarski@psi.ch>
|
|
// SPDX-License-Identifier: GPL-3.0-only
|
|
|
|
#pragma once
|
|
|
|
#include <vector>
|
|
#include <Eigen/Dense>
|
|
|
|
#include "../../common/Coord.h"
|
|
#include "Indexer.h"
|
|
|
|
// The best cells of the fast-feedback indexer for one input cell, before Refine(): three rows per
|
|
// cell (the real-space cell vectors), scores ascending - lower is better. The layout of the GPU
|
|
// library's oCellM()/oScoreV(), so both feed Refine() the same way.
|
|
struct FFBIDXCells {
|
|
Eigen::MatrixX3<float> cells;
|
|
Eigen::VectorX<float> scores;
|
|
};
|
|
|
|
// The fast-feedback indexer on the CPU: same samples, scores and output as the GPU library.
|
|
// cell holds the real-space cell vectors as rows; the first nspots of spots are used.
|
|
FFBIDXCells FFBIDXFindCells(const Eigen::Matrix3f &cell, const std::vector<Coord> &spots, size_t nspots);
|
|
|
|
class FFBIDXIndexerCPU : public Indexer {
|
|
Eigen::Matrix3f input_cell;
|
|
|
|
void SetupUnitCell(const std::optional<UnitCell> &cell) override;
|
|
public:
|
|
std::vector<CrystalLattice> RunInternal(const std::vector<Coord> &coord, size_t nspots) override;
|
|
};
|