main.cpp

/*
 * Copyright (c) 2020, NVIDIA CORPORATION. All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 *  * Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 *  * Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *  * Neither the name of NVIDIA CORPORATION nor the names of its
 *    contributors may be used to endorse or promote products derived
 *    from this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
 * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
 * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR
 * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
 * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
 * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
 * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

#include <cstdio>
#include <vector>

#include <cuda_runtime_api.h>
#include <cublasLt.h>

#include "sample_cublasLt_LtSgemmSimpleAutoTuning.h"
#include "helpers.h"

void printAlgo(const cublasLtMatmulAlgo_t& algo) {
    int algoId, tile, swizzle, customOption, numSplitsK, reductionScheme;

    checkCublasStatus(cublasLtMatmulAlgoConfigGetAttribute(&algo, CUBLASLT_ALGO_CONFIG_ID, &algoId, sizeof(algoId), NULL));
    checkCublasStatus(cublasLtMatmulAlgoConfigGetAttribute(&algo, CUBLASLT_ALGO_CONFIG_TILE_ID, &tile, sizeof(tile), NULL));
    checkCublasStatus(cublasLtMatmulAlgoConfigGetAttribute(&algo, CUBLASLT_ALGO_CONFIG_SPLITK_NUM, &numSplitsK, sizeof(numSplitsK), NULL));
    checkCublasStatus(cublasLtMatmulAlgoConfigGetAttribute(&algo, CUBLASLT_ALGO_CONFIG_REDUCTION_SCHEME, &reductionScheme, sizeof(reductionScheme), NULL));
    checkCublasStatus(cublasLtMatmulAlgoConfigGetAttribute(&algo, CUBLASLT_ALGO_CONFIG_CTA_SWIZZLING, &swizzle, sizeof(swizzle), NULL));
    checkCublasStatus(cublasLtMatmulAlgoConfigGetAttribute(&algo, CUBLASLT_ALGO_CONFIG_CUSTOM_OPTION, &customOption, sizeof(customOption), NULL));

    printf("algo={ Id=%d, tileIdx=%d splitK=%d reduc=%d swizzle=%d custom=%d }\n",
        algoId, tile, numSplitsK, reductionScheme, swizzle, customOption);
}

int main() {
    TestBench<float> props(1024, 1024, 1024, 2.0f, 0.0f, 1024 * 1024 * 4);

    cublasLtMatmulAlgo_t algo;

    props.run([&props, &algo] {
        LtSgemmSimpleAutoTuning(props.ltHandle,
                                CUBLAS_OP_N,
                                CUBLAS_OP_N,
                                props.m,
                                props.n,
                                props.k,
                                &props.alpha,
                                props.Adev,
                                props.m,
                                props.Bdev,
                                props.k,
                                &props.beta,
                                props.Cdev,
                                props.m,
                                props.workspace,
                                props.workspaceSize,
                                algo);
    });

    printAlgo(algo);

    return 0;
}