Struct AMDOpenCLSimulator 

pub struct AMDOpenCLSimulator { /* private fields */ }

AMD GPU-optimized quantum simulator using OpenCL.

In this build there is no OpenCL runtime, so this type cannot be constructed (see AMDOpenCLSimulator::new). The type and its associated kernel-source templates are retained for callers that link a real OpenCL toolchain elsewhere.

Implementations

impl AMDOpenCLSimulator

pub fn new(_config: OpenCLConfig) -> Result<Self>

Create a new AMD OpenCL simulator.

HONEST AVAILABILITY GATE: this build links no OpenCL/AMD ROCm runtime, so AMD device discovery and GPU kernel dispatch are impossible. Constructing a working simulator here would require fabricating device discovery and kernel timings, so we fail loudly instead.

pub const fn get_device_info(&self) -> Option<&OpenCLDevice>

Get device information (only present with a real runtime).

pub const fn get_kernels(&self) -> &HashMap<String, OpenCLKernel>

Get the compiled kernel-source templates.

pub const fn get_stats(&self) -> &OpenCLStats

Get performance statistics.

pub const fn config(&self) -> &OpenCLConfig

Get the backend configuration.

pub fn kernel_source_templates( config: &OpenCLConfig, ) -> HashMap<String, OpenCLKernel>

Build the standard set of quantum OpenCL kernel source templates.

Returns the kernels as text only; nothing here compiles or executes on a device. Exposed as an associated function so the templates remain reachable/inspectable even though Self::new cannot succeed in this build.

pub fn build_options_for(config: &OpenCLConfig) -> String

Build OpenCL kernel-compilation options for the given configuration.