Redefining Instrumentation Architectures – Technologies for Mixed-Signal, Streaming Applications

What do Google, Wikipedia, and YouTube have in common?

These are all Web-based applications that have changed the Web from a producer-consumer model, in which a relatively small number of people produce Web content, to a peer-to-peer model, in which each user receives highly customized content. This new paradigm of Internet applications has been dubbed Web 2.0.

At the same time, our world has become increasingly software-oriented.
The features of devices we use every day such as smart phones, set-top boxes, and even automobiles are increasingly defined by the embedded software in these devices.

Instrumentation can take advantage of these two macro trends – user-defined customization and increased software focus – so you can completely customize your application. The concept of user-defined instrumentation is not new; in fact, it has been around for more than two decades in the form of virtual instrumentation. However, the technologies driving these trends now have matured to create a tipping point toward this new software-defined model. Similar to the Web, the difference is distinct enough to be called Instrumentation 2.0.

Two key technologies are driving this change – the high-speed PCI Express bus and multicore processing. Now you can address applications that once required proprietary, custom solutions with new products on the cutting edge of Instrumentation 2.0.

PCI Express Increases Data Bandwidth

With the software-oriented, user-defined instrumentation approach, the focus is now on the PC and its associated technologies. Because you send signals acquired with I/O hardware to the PC over a data bus, it is critical for the bus technology to keep up with the ever-increasing I/O resolution and speed. PCI Express delivers on this need for faster data throughput with the highest throughput of any other commercial communications bus.

PCI Express, available in x1, x4, x8, and x16 links (pronounced “by 1,” “by 4,” and so on), provides 250 MB/s of throughput per lane with very low latency. The x1 and x4 options, most commonly used for instrument-class hardware, provide 250 MB/s and 1 GB/s (four lanes at 250 MB/s) of dedicated throughput, respectively. The x16 link, which provides 4 GB/s of throughput, is now commonly used in new PCs for plug-in video cards.

Multicore Technology Improves Processor Performance

While you can transfer signals from high-speed I/O hardware to the PC with the increased throughput of PCI Express, the PC must have enough processing power to process all of the data transferred on the bus. The latest advancement in processor technology is to place multiple cores, or computing engines, within a single processor. Currently, both Intel and AMD have released dual-core processors, and future processors will expand the number of cores to four or more. In fact, Intel is targeting to deliver an 80-core processor in a mere five-year period.

Multitasking operating systems such as Windows XP and Windows Vista, as well as multithreaded applications such as National Instruments LabVIEW software, can take full advantage of the new parallel processing capability introduced by multicore technology. Writing multithreaded applications in text-based programming languages, such as C, is nontrivial and requires expertise in the semantics of creating and managing the threads and passing data between them in a thread-safe way. On the other hand, with NI LabVIEW, you can readily take full advantage of the additional horsepower provided by multicore processors because the LabVIEW graphical environment naturally lends itself to parallel programming. In LabVIEW, two loops that do not share a data dependency automatically execute in separate threads.

Consider Table 1 for a representative example of the performance improvement achieved with an existing LabVIEW application on a dual-core processor versus a single-core processor.

[+] Enlarge Image

Table 1. By executing tasks in parallel, the same LabVIEW application, which finds all the prime numbers in the first 1,000,000 natural numbers, runs 47.74 percent faster on a dual-core processor.

PCI Express and Multicore Combine in the PXI Platform

The PXI industry standard was introduced more than 10 years ago, and today, engineers are using PXI in tens of thousands of applications. Additionally, the PXI Systems Alliance (PXISA), which governs the PXI standard and includes more than 70 members, continues to introduce hundreds of new and innovative PXI products. With the introduction of PCI Express, the PXISA quickly adopted this higher-speed data bus in the PXI Express specification, and new PXI Express products began rolling out on the market.

For example, National Instruments has already released PXI Express chassis, controllers, and modules. The new NI PXIe-1065 18-slot chassis features seven hybrid slots that work with either PXI Express or PXI modules. The NI PXIe-8106 dual-core controller features a 2.16 GHz Intel Core 2 Duo T7400 dual-core processor.

The new NI PXIe-5122 digitizer and NI PXIe-6537/36 digital I/O modules provide a solution for high-speed data recording/playback and mixed-signal applications. You can stream your signals at the full data rate (400 MB/s for NI PXIe-5122 and 200 MB/s for NI PXIe-6537) to the processor over the PCI Express bus in the backplane of your PXI chassis. You can then process your signals on the processor and/or stream these signals to PC memory or to an external array of hard disks.

By combining the high data throughput of PXI Express with powerful, new PXI controllers based on multicore technology, PXI systems deliver an Instrumentation 2.0 architecture with one industry-standard platform.

PXI Express Solves Challenging Application Problems

While you can stream your signals at unprecedented data rates with the new digitizer and digital I/O modules, your system’s measurement capabilities are not limited to only these instruments. Because PXI Express provides full compatibility with existing PXI modules, you can build a complete mixed-signal test solution. These applications include high-speed imaging, signal intelligence, RF/IF data streaming, digital video generation/streaming from PC, and more. Below are two specific examples that benefit from the higher-performance PXI Express platform.

 IF Data Streaming – High-Speed Recording/Playback

Signal intelligence applications often require the ability to stream intermediate frequency (IF) signals to disk. Using the new NI PXIe-5122 two-channel, 100 MS/s, 100 MHz, 14-bit digitizer with the new Conduant PXIe-416 StreamStor solution, you can acquire and stream the IF signal directly to hard disk, bypassing the digitizer’s onboard memory and
processor. With Conduant’s StreamStor solution, you can connect up to four NI PXIe-5122 modules, each with a dedicated 400 MB/s streaming throughput, and stream the data to four hard drives, each with 8 TB of  storage. With this scheme, you can stream a combined 1.2 GB/s continuously for 5.8 hours.

If streaming to hard disk is not required in your application, you can stream the data from the digitizer over the PCI Express bus at the full 400 MB/s to the processor. If you were using a multicore processor such as the NI PXIe-8106, you can leverage the new processing power by implementing a digital downconverter and other analysis on the two cores of the processor. Because of the parallel and multithreaded approach of
LabVIEW, your application readily harnesses the full computing power of the two cores.

[+] Enlarge Image

Figure 1. You can achieve IF data streaming with the NI PXIe-5122 100 MS/s, 14-bit digitizer and Conduant PXIe-416 StreamStor modules. Using three antennas connected to three digitizers, you can use triangulation algorithms to determine the location of the target.

Stream Digital Patterns from PC Memory or Hard Disk

Many high-speed digital applications such as interfacing to memory chips, emulating custom protocols, and testing image sensors and display panels require custom signal streaming from the PC memory or disk to the device under test (DUT). Until the introduction of NI PCIe-6537/36, the onboard memory of the digital I/O module was used to store the digital waveform/ pattern and play the signal from the module’s onboard memory. The onboard memory on the module not only makes the test system more expensive but also limits the length of the test pattern.

[+] Enlarge Image

Figure 2. This diagram illustrates LCD testing with streaming digital video generation
using the NI PXIe-6537 50 MHz DIO module.

The NI PXIe-6537/36 and NI PCIe-6537/36 digital I/O devices solve these issues so you can stream your test pattern/waveform straight from the PC memory or disk over the PCI Express bus. These modules use a x1 PCI Express link, allowing a sustainable data rate of 200 MB/s, which is the maximum data rate of the module. In the example shown in Figure 2, the NI PXIe-6537 module streams the digital video pattern from the PC memory at a full data rate of 200 MB/s over the PCI Express bus to the display unit, an LCD DUT, for testing the video quality of the LCD.

Customize Your Application with Instrumentation 2.0

The PCI Express bus and multicore processors are two of the latest PC technologies that enable Instrumentation 2.0, which gives you the ability to completely customize your application in software. PXI and LabVIEW, which inherently take advantage of multicore technology, provide a single platform for upgrading your test system with the latest PC technologies. Using PXI, you now can meet application challenges that were only solvable using the previous generation of expensive, proprietary test systems.

A Closer Look at Instrumentation 2.0

The Traditional Approach – Instrumentation 1.0

[+] Enlarge Image

With traditional, stand-alone instruments, if signal bandwidth is greater than bus bandwidth (BWsignal > BWbus), you need expensive onboard memory and/ or onboard processing. This is the reason why measurements are implemented in firmware by the vendor inside the box in the traditional instrumentation approach. Only the final measurements, as defined by the vendor, are sent over to the PC, and you generally do not receive the actual data for signal processing or custom measurements. The increasing need for user-defined test systems calls for a new approach to instrumentation.

Instrumentation Architecture Redefined

[+] Enlarge Image

The high-throughput PCI Express bus creates a shift in the instrumentation paradigm. For most instruments, bus bandwidth exceeds signal bandwidth (BWbus > BWsignal), so you can send the entire waveform over the PCI Express bus. Of course, the processor must be able to keep up with all the incoming data, and multicore processors promise to deliver the processing power needed for an increasingly broad set of applications. Such a method obviates the need for onboard signal processing and onboard memory in most applications. The real power of this redefined architecture lies in the fact that you now can completely customize your application, in addition to getting the final results.

[+] Enlarge Image

New PXI Express Products

NI PXIe-5122 Digitizer

• 100 MS/s, 100 MHz, 14-bit, twosimultaneous channels
• Streaming at a full rate of 400 MB/s (dedicated throughput)
• Synchronization of multiple modulesto within picoseconds

NI PXIe-6537/36 Digital Waveform Generator/Analyzers

• 25 and 50 MHz digital input and digital output
• 32 channels with per-channel directional control
• Streaming at up to 200 MB/s (dedicated throughput)

NI PXIe-8106, PXI-8106, PXI-8106 RT Embedded Controllers

• 2.16 GHz Intel Core 2 Duo T7400 dual-core processor
• Up to 2 GB DDR2 RAM
• Highest-performance PXI Express embedded controller

This article first appeared in the Q2 2007 issue of Instrumentation Newsletter.

Reader Comments | Submit a comment »

Vision advantages
What are the expected advantages of the Intrumentation 2.0 over an instrumentation 1.0 vision application?
- Jean Depéry, Micromag. depery@micromag.ch - Jun 6, 2007