A: The number of systems that can be daisy chained is system-dependent, and is influenced by the host OS/BIOS and the PCI Express host architecture. Based on testing, you should be able to daisy chain four to seven chassis, depending on your system. Each string of daisy-chained chassis will be limited to the bandwidth of a x1 PCI Express lane (250 MB/s theoretical).
See the Developer Zone Tutorial: System Configuration Effects on NI MXI-Express RIO Daisy Chain Depth for more details.
A: Yes, but not natively. The system clock is not exported on the MXI-Express bus. To synchronize multiple chassis, use digital input/output (DIO) modules and the field-programmable gate array (FPGA) to create a system-wide clock.
A: The maximum number of chassis is OS/BIOS-dependent, and depends on the number of PCI Express bus segments in the system. A single device consumes multiple segments (MXI-Express RIO uses four whereas MXI-Express host cards use three). The maximum number of bus segments allowed by the PCI Express spec is 255, although some PC vendors make assumptions in their BIOS to limit the number based on its design assumptions.
Assuming the PC BIOS supports up to 255 bus segments, and that the controller takes about 30 bus numbers for other devices, the maximum recommended configuration is eight daisy-chained lines of six chassis each, for a total of eight MXI adapters and 48 MXI-Express RIO chassis. This configuration should use (48 MXI-Express RIO) * (4 bus segments per MXI-Express RIO) + (8 chassis) * (3 bus segments per MXI-Express host card) + (30 reserved bus numbers) = 246 of the 255 PCI Express bus numbers.
A: MXI-Express RIO is compatible with the following:
A: No, the NI PXIe-8360 X1 MXI-Express for PXI Express interface is a slave-only device, meaning you cannot use it from a PXI host controller to connect to a MXI-Express chassis. To use MXI-Express RIO with your PXI or PXI Express chassis, you need to use the NI ExpressCard-8360 or NI PXI/PXIe-8364.
A: You have to calculate the maximum data throughput rates from each module in the system, as well as take into account which chassis in the daisy chain contains the modules. Attached below is a system configurator spreadsheet, which you can use to calculate maximum allowable throughput numbers for a specific system.
Note: This spreadsheet gives you a good approximation, but was calculated based on benchmarks from a specific hardware configuration. Actual throughput numbers from your system are dependent on the host OS, BIOS, and software architecture used to stream data.
A: With MXI-Express RIO, you can have up to three DMA FIFOs per chassis in a daisy chain. So, if you have three chassis daisy chained together to a host, you can have three DMA FIFOs per chassis for a daisy chain total of nine DMA FIFOs.
A: The MXI-Express RIO device and any downstream devices can detect that there has been a communication loss with the host and chose to either power down or continue to run LabVIEW FPGA code. If a MXI-Express RIO device loses power in the chain, any devices downstream of it would detect that the host went away. In either case, the upstream devices will remain connected provided the host PC does not blue screen from the unplugging/power-down event.
A: The max cable length is seven meters between each chassis. For a longer distance distributed solution, use the NI EtherCAT RIO or NI Ethernet RIO expansion chassis, which allow you to reach distances of up to 100 meters before a hub or repeater. For even further distances, consider NI wireless sensor networks (WSNs), which allow you to reach distances up to 300 meters between nodes.
A: No, the FPGA must be accessed and programmed with the NI LabVIEW FPGA Module.
A: No, this has been a common point of feedback and is under review.