Jason Zavaglia jpz

## dynamicParallelismCMAKE.md

      
              1 file
            
          
              1 fork
            
          
                0 comments
              
            
              3 stars
            
          
                srivathsanmurali
                / dynamicParallelismCMAKE.md
            
            
              Last active
              August 12, 2024 15:47
            
              
                Compileing CUDA with dynamic parllelism using CMAKE
              
          
    Dynamic parallelism using CUDA and CMAKE

Requirements:


compute_35 or higher cards. I am using Pascal (compute_61)

CMAKE MAGIC


set(CUDA_SEPARABLE_COMPILATION TRUE)
list(APPEND CUDA_NVCC_FLAGS -gencode arch=compute_61,code=sm_51;)
target_link_libraries(yourCudaLib ${CUDA_cudadevrt_LIBRARY}


## proxy.go
package main

import (
	"bytes"
	"encoding/hex"
	"flag"
	"fmt"
	"io"
	"log"
	"net"
	package main

	import (
	"bytes"
	"encoding/hex"
	"flag"
	"fmt"
	"io"
	"log"
	"net"