diff --git a/libc/utils/gpu/loader/nvptx/Loader.cpp b/libc/utils/gpu/loader/nvptx/Loader.cpp
--- a/libc/utils/gpu/loader/nvptx/Loader.cpp
+++ b/libc/utils/gpu/loader/nvptx/Loader.cpp
@@ -190,6 +190,7 @@
           // Wait until the memory allocation is complete.
           while (cuStreamQuery(memory_stream) == CUDA_ERROR_NOT_READY)
             ;
+          buffer->data[0] = static_cast<uintptr_t>(dev_ptr);
         };
         rpc_recv_and_send(port, malloc_handler, data);
       },