Considerations for Porting Applications to 64-bit
Before you can compile a 64-bit project, you must update your code so that it is 64-bit ready. Use the guidelines recommended by National Instruments to prepare your code so that you can use one code base for both 32-bit applications and 64-bit applications.
Although you can run 32-bit applications on 64-bit systems, 32-bit applications do not fully take advantage of the capabilities of 64-bit platforms.
Consider porting your application if you need the additional processing power available with a 64-bit application. For example, you might want to port your application if your project handles data larger than 2 GB.
Remember that converting your application to a 64-bit application involves updating your code. In addition, consider the following before you port your application:
- The following LabWindows/CVI components are not supported in 64-bit applications:
- DataSocket Library (dataskt.fp), including data binding functionality for user interface controls
- Scroll bar custom control (scroll.fp)
- Easy tab custom control (easytab.fp)
- OpenGL control (cviogl.fp)
- LabVIEW RT board interface control (cvirtsup.fp)
- 3D Graph control (3dgraphctrl.fp)
- StartPCSound (toolbox.fp)
- StopPCSound (toolbox.fp)
- LabWindows/CVI low-level support driver
- Functions in the External Modules class in the Utility Library
- Function panel customization (fpcustom.fp) instrument driver
- National Instruments Device Driver libraries might not include support for 64-bit applications. Refer to the latest version of the product documentation at ni.com/manuals for the most up-to-date information regarding 64-bit support for LabWindows/CVI for device driver libraries.
- You must create a separate distribution for 64-bit applications. To create a 64-bit distribution, select Type: Windows Installer (.msi) or Package and Supported OS: Windows 64-bit only in the New Distribution dialog box.
- If you have the LabWindows/CVI environment installed on a 32-bit computer and you want to debug your 64-bit application, you must use remote debugging to debug your application as it runs on a 64-bit computer.
- When you create a project containing Analysis or Advanced Analysis Library functions, the numerical results returned by the functions are slightly different between the 32-bit version of the application and the 64-bit version of the application. The numerical results differ after approximately 9-12 digits.
When you use DBL_MIN or DBL_MAX values, 64-bit applications running on 64-bit OSes might return values as infinity, but 32-bit applications might return values as numbers. This behavior occurs because floating point operations, when running as part of a 32-bit application, use 80-bit wide Floating Point Unit (FPU) registers for temporary values; however, floating point operations, when running as part of a 64-bit application, use 64-bit wide FPU registers. For more information, refer to ni.com/info and enter exv9pb as the Info Code.