Software-Timed Generations

With a software-timed generation, software controls the rate at which data is generated. Software sends a separate command to the hardware to initiate each update. In NI-DAQmx, software-timed generations are referred to as on-demand timing. Software-timed generations are also referred to as immediate or static operations. They are typically used for writing a single value out, such as a constant digital value.

Hardware-Timed Generations

With a hardware-timed generation, a digital hardware signal controls the rate of the generation. This signal can be generated internally on your device or provided externally.

Hardware-timed generations have several advantages over software-timed generations:

• The time between samples can be much shorter.
• The timing between samples can be deterministic.
• Hardware-timed acquisitions can use hardware triggering.

PCIe-6321 and PXIe-6321 devices support buffered hardware-timed operations and hardware-timed single point operations.

A buffer is a temporary storage in computer memory for to-be-transferred samples. Data is moved from a PC buffer to the DAQ device’s onboard FIFO using DMA before it is written to the output lines one sample at a time. Buffered generation typically allow for much faster transfer rates than non-buffered acquisitions because data is moved in large blocks, rather than one point at a time.

One property of buffered I/O operations is the sample mode. The sample mode can be either finite or continuous:

• Finite sample mode acquisition refers to the acquisition of a specific, predetermined number of data samples. Once the specified number of samples has been written out, the generation stops.
• Continuous generation refers to the generation of an unspecified number of samples. Instead of generating a set number of data samples and stopping, a continuous generation continues until you stop the operation. There are several different methods of continuous generation that control what data is written. These methods are regeneration, FIFO regeneration and non-regeneration modes:
  • Regeneration is the repetition of the data that is already in the buffer. Standard regeneration is when data from the PC buffer is continually downloaded to the FIFO to be written out. New data can be written to the PC buffer at any time without disrupting the output. Use the NI-DAQmx write property regenMode to allow (or not allow) regeneration. The NI-DAQmx default is to allow regeneration.
  • With FIFO regeneration, the entire buffer is downloaded to the FIFO and regenerated from there. Once the data is downloaded, new data cannot be written to the FIFO. To use FIFO regeneration, the entire buffer must fit within the FIFO size. The advantage of using FIFO regeneration is that it does not require communication with the main host memory once the operation is started, thereby preventing any problems that may occur due to excessive bus traffic. Use the NI-DAQmx UseOnlyOnBoardMemeory DO channel property to enable or disable FIFO regeneration.
  • With non-regeneration, old data is not repeated. New data must be continually written to the buffer. If the program does not write new data to the buffer at a fast enough rate to keep up with the generation, the buffer underflows and causes an error.

Hardware-timed single point (HWTSP) operations are typically used to write single samples at known time intervals. While buffered operations are optimized for high throughput, HWTSP operations are optimized for low latency and low jitter. In addition, HWTSP can notify software if it falls behind hardware. These features make HWTSP ideal for real time control applications. HWTSP operations, in conjunction with the wait for next sample clock function, provide tight synchronization between the software layer and the hardware layer. Refer to the NI-DAQmx Hardware-Timed Single Point Lateness Checking document for more information. To access this document, go to ni.com/info and enter the Info Code daqhwtsp.