`nvidia` target: simulation errors when setting `CUDAQ_MAX_CPU_MEMORY_GB` beyond system memory capacity

Required prerequisites

• Consult the security policy. If reporting a security vulnerability, do not report the bug using this form. Use the process described in the policy to report the issue.
• Make sure you've read the documentation. Your issue may be addressed there.
• Search the issue tracker to verify that this hasn't already been reported. +1 or comment there if it has.
• If possible, make a PR with a failing test to give us a starting point to work on!

Describe the bug

Using the CUDAQ_MAX_CPU_MEMORY_GB environment variable setting, when an unlimited setting (None) is used or a value larger than the actual physical system's memory available, the simulation may fail (e.g., with a cuda error) or produce invalid simulation results.

Steps to reproduce the bug

• Run a simulation that requires more qubits than GPU memory.

• Set CUDAQ_MAX_CPU_MEMORY_GB to a value larger than the physical host memory.

Expected behavior

The simulation should work as expected or at least a more descriptive error message should be displayed.

Is this a regression? If it is, put the last known working version (or commit) here.

Not a regression

Environment

• CUDA Quantum version: 0.8
• Python version: 3.10
• C++ compiler: gcc-11
• Operating system: Ubuntu 22.04

Suggestions

The current workaround is to set CUDAQ_MAX_CPU_MEMORY_GB to a valid value, taking into account the physical system memory configuration.

No source files, tests, or entry points are named. Start by reproducing the nvidia target simulation with CUDAQ_MAX_CPU_MEMORY_GB set to None and above physical memory, then trace its environment-variable handling. Done means the simulation behaves safely or emits a descriptive error for invalid memory limits.