STS. Smarter devices. Smarter test systems.
- 1. NI Automated Test Outlook 2016 Smarter devices. Smarter test systems.
- 2. Table of Contents 05 Harvesting Production Test Data Semiconductor organizations pioneer real-time data analytics to reduce manufacturing test cost. 07 Life-Cycle Management Is All About Software Obsolescence, OS churn, and compatibility challenge long-life-cycle projects—an age-old problem warrants revisiting. 09 The Rise of Test Management Software Off-the-shelf test executives are effective solutions for the influx of new programming languages. 11 Standardizing Platforms From Characterization to Production RFIC companies employ IP reuse and hardware standardization across the product design cycle to reduce cost and decrease time to market. 13 Making (mm)Waves in Test Strategy Test managers are adopting modular solutions to economically validate high-frequency components.A Technology and Business Partner Since 1976, companies around the world including BMW, Lockheed Martin, and Sony have relied on NI products and services to build sophisticated automated test and measurement systems. Test delivers value to your organization by catching defects and collecting the data to improve a design or process. Driving innovation within test through technology insertion and best-practice methodologies can generate large efficiency gains and cost reductions. The goal of the Automated Test Outlook is to both broaden and deepen the scope of these existing efforts and provide information you need to make key technical and business decisions.
- 3. Organizational Test Integration OptimizingTest Organizations Test Economics Organizational Proficiency Testing the Hybrid Approach mmWaveTest Strategies Business Strategy Business Strategy ArchitectureArchitecture ComputingComputing SoftwareSoftware I/OI/O System Software Stack Measurements and Simulation in the Design Flow Software-Centric Ecosystems ManagedTest Systems Bridging the Big Data Chasm Rise ofTest Management Software Heterogeneous Computing PCI Express External Interfaces Big Analog Data Cloud Computing forTest The Core of the Matter Harvesting ProductionTest Data IP to the Pin Proliferation of Mobile Devices Test Software Quality ScalableTest Software Architecture Driven by Necessity Life-Cycle Management Portable Measurement Algorithms Moore’s Law Meets RFTest Redefining the Notion of Sensors From 1G to 5G Characterization to Production 2011 2012 2013 2014 2015 2016 AUTOMATED TEST OUTLOOK 2016 How We Arrived at the Trends As a supplier of test technology to more than 35,000 companies each year, we receive a broad range of feedback across industries and geographies.This broad base creates a wealth of quantitative and qualitative data to draw on. We stay up to date on technology trends through our internal research and development activities. As a technology-driven company, we invest more than 16 percent of our annual revenue in RD. But as a company that focuses on moving commercial technology into the test and measurement industry, our RD investment is leveraged many times over in the commercial technologies we adopt. Thus, we maintain close, strategic relationships with our suppliers. We conduct biannual technology exchanges with key suppliers that build PC technologies, data converters, and software components to get their outlook on upcoming technologies and the ways these suppliers are investing their research dollars. Then we integrate this with our own outlook. We also have an aggressive academic program that includes sponsored research across all engineering disciplines at universities around the world. These projects offer further insight into technology directions often far ahead of commercialization. And, finally, we facilitate advisory councils each year for which we bring together leaders from test engineering departments to discuss trends and share best practices. These councils include representatives from every major industry and application area—from fighter jets to the latest smartphone to implantable medical devices. The first of these forums, the Automated Test Customer Advisory Board, has a global focus and is in its 16th year. We also conduct meetings, called regional advisory councils, around the world. Annually, these events include over 300 of the top thought leaders developing automated test systems. We’ve organized this outlook into five categories (see above figure). In each of these categories, we highlight a major trend that we believe will significantly influence automated test in the coming one to three years. We update the trends in these categories each year to reflect changes in technology or other market dynamics. We will even switch categories if the changes are significant enough to warrant it. As with our face-to-face conversations on these trends, we hope that the AutomatedTest Outlook will be a two-way discussion. We’d like to hear your thoughts on the industry’s technology changes so we can continue to integrate your feedback into this outlook as it evolves each year. Email ato@ni.com or visit ni.com/test-trends to discuss these trends with your peers.
- 4. OPTIMAL+ BIG DATA INFRASTRUCTURE USING NI STS Optimal+ Portal MANUFACTURING FACILITIES NI STS NI STS NI STS Proxy Server Test Data Commands and Control The explosion of Internet of Things devices will further challenge companies’ capabilities: products are becoming more complex while quality expectations soar, which causes test programs and the data they generate to increase exponentially in size; data is collected from geographically dispersed sources and processes; and RMA prevention requires the retention of data for months and even years. However, all is not lost. Leading semiconductor companies are pioneering a multifaceted big data analytics strategy to improve product yields, prevent escapes, and streamline RMA management. This groundwork is creating a blueprint for how manufacturing test organizations in other industries can realize the benefits of data analytics. Collection and Detection Providing a solid foundation for big data analytics in manufacturing operations presents two primary challenges. The first is collecting data from the tester in a standardized format and quickly delivering it, regardless of where that tester is located. Most semiconductor vendors have supply chains composed of separate companies in different countries conducting manufacturing, assembly, and test. This often prevents product teams from gaining rapid access to clean, consistent test data. A standardized approach to collecting data across a global supply chain is necessary for companies to best leverage the knowledge contained within that data. The second challenge is storing data for long periods of time and ensuring it is readily accessible for detailed analysis. For companies involved in market segments that demand high-quality electronic products, processes that can help limit test escapes, rapidly resolve field failures, and manage RMAs are critical. These processes have generated a greater need for engineering operations to work closely with IT to plan how global test data is managed. Will the data be stored on-site, in the cloud, or some combination of the two? Big data solutions need to store and provide rapid access to enormous quantities of information (tens to hundreds of terabytes) and help teams accomplish tasks in the most efficient way possible. From Reactive to Proactive Access to data in real time helps solve many problems but only if a user can quickly mine that data. Data by itself is meaningless until it is analyzed. And in the case of semiconductor manufacturing data, much of that value is time-sensitive. The faster a user can analyze data, the more valuable the analysis results. In many industries, data mining is a “reactive” process. Something goes wrong (for example, too many field failures), so a product team starts to analyze data to determine the problem, correct it, and then put steps in place to keep it from happening again. But in many cases, the damage is done, either in terms of a product recall or a negative impact on product revenue or even the market valuation of a company. The challenge is catching these problems early enough to mitigate them. One of the major benefits of big data analytics is being “proactive” in automatically analyzing, in minutes, any amount of data 24 hours a day. In semiconductor manufacturing, engineers can embed their knowledge into automated rules engines that mine data 24/7, search for manufacturing issues, and provide immediate alerts. Acting on Data Eliminating the “find a needle in a haystack” problem is a paradigm shift for product and yield engineers. It helps emphasize acting on issues instead of continually looking for problems and finding them too late to effect meaningful change. A simple example is a probe card or load board that is starting to fail. The immediate effect is usually a drop in yield. But how long will it take for someone to notice a problem? In some instances, the problem could go unnoticed for hours or more. If yield drops 6 percent during that time, any material tested during that timeframe is irretrievably lost. This is where big data analytics comes into play. Yield engineers can set up a rule that checks for any statistically relevant drop in yield at any test site in the entire supply chain. As soon as the rule is triggered, the rule engine alerts the appropriate people to take action within minutes and, in this case, preserve yield entitlement. This also helps improve quality. One issue that affects semiconductor device quality is the number of“touchdowns” during wafer sort. Depending on the retest policy, in the desire to “pass” a device, a die may endure too many touchdowns, which puts its long-term quality and reliability at risk. But if the device tests as “good,” what can be done to prevent it from being shipped into the supply chain? Using analytics, every die from every lot can be evaluated between wafer sort and final test to see how many touchdowns it receives. If any given die records more touchdowns than is deemed acceptable, that die can be rebinned as “bad” before it goes to assembly or final test, which removes a highly suspect die from the supply chain. From Pain Comes Progress Whatever a company’s pain points may be—yield, quality, or productivity—big data solutions can help improve operational metrics by automatically analyzing manufacturing data based on rules that embody the complete knowledge and expertise of its operations teams. Today, many of the world’s largest semiconductor companies, both IDM and fabless, are leveraging the power of big data analytics to help them collect, detect, and act on global manufacturing data and drive yield and productivity while improving overall quality. This ultimately increases profit margins and market share. Harvesting Production Test Data Every year, semiconductor companies accumulate tens of terabytes of manufacturing test data whose inherent value and bottom line impact are lost. And it’s only getting worse. “Companies like Optimal+ are delivering on the promise of big data for semiconductor manufacturing by providing near real-time ability to analyze and act on the insights in the data, lowering the cost of test and improving the product quality. This trend will only continue as other industries start to embrace the benefits of big data.” —Mike Santori, Business and Technology Fellow, NI ni.com/ato 05 AUTOMATED TEST OUTLOOK 2016 Article contributed by: David Park, Vice President of Worldwide Marketing, Optimal+
- 5. 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 0 400 800 1200 1600 2000 REVENUE(MILLIONSUSD) PXI REVENUE VXI REVENUE AXIe REVENUE INNOVATION AND STABILITY MAKE PXI THE STANDARD PLATFORM FOR AUTOMATED TEST simply opening, saving, and revalidating a test program set (TPS), or test sequence, can cost hundreds of thousands of dollars. This has created an environment where companies must rethink their software strategies or risk hemorrhaging money to sustain legacy testers. Since minor software changes can greatly impact TPS compatibility, instrument vendors should offer TPS-compatible hardware migration options. This includes preserving driver functionality, APIs, and dependencies between driver versions to minimize the impact on the hardware abstraction layer. For example, NI is collaborating with Astronics Corporation to bring remaining VXI instruments into the PXI platform, such as the Astronics PXIe-2461 frequency time interval counter, which preserves TPS compatibility with legacy systems. Despite their best efforts, vendors cannot always provide TPS-compatible alternatives. In these situations, a common approach is emulating legacy instrument functionality. Recently, engineers have adopted software-designed instruments with user- programmable FPGAs to augment standard instrument capabilities with custom functionality to emulate legacy behavior. For example, filters and triggers that were common in instruments 20 years ago and obsolete in today’s instruments can be reengineered. Coming Full Circle Whether you’re managing the B-52 bomber platform or introducing a new line of infotainment systems for the connected car, life-cycle management is critical. It can be either an expensive afterthought or a competitive advantage. In the face of market dominance of mobile technologies, the accelerated decay of legacy instrumentation, and the rising costs of software validation, scalable test architectures and strategies will distinguish best-in-class organizations. One of the largest operational costs associated with automated test systems, especially in the aerospace and defense industry, is the support and maintenance cost over the life of the system. Proactive life-cycle management requires designing maintainable testers, diligently monitoring automated test equipment (ATE), and tracking instrument and component end-of-life (EOL) notifications. While life-cycle management might not be a novel concept, the reality is that the evolution of mobile technology, accelerated hardware obsolescence, and sheer volume of test software are making this task increasingly difficult. Best-in-class organizations are rearchitecting test strategies to gain a competitive advantage amid the growing challenge of life- cycle management. Evolution of OS Life Cycles Within a decade, OS providers have transitioned from releasing a single OS and maintaining it for several years, such as Microsoft Windows XP (which was supported for 13 years), to today’s paradigm that targets mobile users that expect constant upgrades. This requires OS providers to frantically release new versions and retroactively fix bugs in daily updates. Global market intelligence firm IDC forecasts that smartphones and tablets will control 88.4 percent of the smart-connected device market by 2019, leaving portable and desktop PCs with only 11.6 percent. As mobile devices control vast majority of the market, OS providers will continue to prioritize the mobile user. This shift poses a monumental hurdle for test systems that rely on a stable OS to eliminate the need for system revalidation. As a result, some organizations are moving to Linux-based systems to have more control over the OS. Another approach is to minimize the number of OSs to reduce the burden for test engineering and IT organizations. Many legacy test systems contain several OSs (one for each unique box instrument), which introduces the risk of revalidation due to individual OS updates. One major benefit of modular platforms, such as VXI or PXI, is the single OS controlling all instruments in the chassis or system. Accelerated Decay of VXI and Legacy Instruments In the late 1980s and early 1990s, the aerospace and defense community standardized on VXI as the modular commercial off-the-shelf platform for ATE systems. However, as VXI grows obsolete and support diminishes for legacy instruments, programs are under increased pressure to migrate to a stable alternative. This is compounded by a looming RoHS conversion deadline, which will increase the rate of component and instrument EOLs. Over the past decade, PXI has replaced VXI as the de facto modular platform for ATE systems due to the size, performance, cost, and level of innovation in the platform. Global consulting firm Frost Sullivan expects PXI to grow by 17.6 percent annually, which accounts for most of the expected growth for the test and measurement industry. With nearly 70 vendors offering more than 1,500 PXI instruments and a steady stream of innovation, PXI will continue to provide increased value to long-life-cycle ATE systems. TPS-Compatible Migration Paths As teams migrate from VXI-based to PXI-based test systems, the investment required to modernize hardware will typically pale in comparison to that of updating and revalidating software. Due to the criticality of the system and the tight regulations for requirements tracking and software validation, Life-Cycle Management Is All About Software In 2015, the US Department of Defense announced that the B-52 bomber, originally introduced in 1952, will be in operation until 2044—a life cycle of nearly 100 years. “The cost to rewrite a TPS due to the replacement of legacy/obsolete instrumentation in a test system is approximately $150k/TPS. When multiplied across dozens of TPS per test system and three to five generations of test equipment over the life of a test system, the potential savings in TPS costs alone are very significant—any efforts that vendors can make to smooth this transition will prove to be invaluable.” —David R. Carey, PhD, Associate Professor of Electrical Engineering, Wilkes University Source: Frost Sullivan ni.com/ato 07 AUTOMATED TEST OUTLOOK 2016
- 6. HETEROGENEOUS TEST DEVELOPMENT APPROACHES Python C LabVIEW .NET FORTRAN C VS. Python LabVIEW .NET FORTRAN TEST MANAGEMENT SOFTWARE organizational whiplash. This can be a costly exercise that often requires code migration, revalidation of the codebase, and additional training in the new language. The best test managers must look to a newer approach to test-system development that builds a heterogeneous system out of multiple languages. This kind of approach allows a team to use multiple languages, each for its own advantages, to build more powerful test systems. For instance, Python could be used for scripting validation and verification tests based on code developed by RD engineers. In the same system, C# could be used to develop an object- oriented interface for custom hardware or to existing .NET libraries while LabVIEW communicates with and gathers data from hardware. Since all languages are designed to tackle specific applications, using each to its strengths would ultimately save time and money. Though beneficial, this approach can present a new challenge to test system development: different languages now need to work and communicate together to form a single system. To resolve this, all test engineers need to understand not just the one environment they specialize in but also the others to adequately interface with them. The Software Solution Test departments are now turning to commercial off-the-shelf test management software to act as a Rosetta Stone of sorts between different languages. This software not only offers users a common environment in which they can work with any type of test code but also completes executive tasks, such as sequencing and calling each test, handling data logging, and generating reports. Each engineer can then focus on writing the best test for each component of the DUT without needing to worry about how to communicate with the other portions of code. And since engineers can use the environments they are most comfortable with, an organization can focus on hiring engineers with skills unique to its applications, even if they don’t have prior knowledge of, or training in, the company’s required language. Additionally, test managers can take advantage of the full power of a heterogeneous design while avoiding the new challenges that such a design introduces. This includes the use of test management software for a more modular development process, which produces a system that’s easier to maintain and upgrade because each component can be updated individually without affecting the rest of the test system. Ultimately, test executives typically include the support of commercial vendors that continually patch and upgrade the test software, which further offsets the cost of maintenance and increases the sustainability of these systems. These advantages, combined with the benefits of a heterogeneous test system design, are how the best test managers are building the future of automated testing. So how did we get here? The evolution is rooted deeply in the history of programming languages and the thirst for higher levels of abstraction, and it starts with the first high-level programming language, FORTRAN. Developed in 1953 by John Backus, FORTRAN addressed the need for a greater level of abstraction for machine processes built around the way that humans naturally communicate their ideas—through language. After the success of FORTRAN, more languages like C, Pascal, ATLAS, and PAWS were developed, each bringing new constructs and models of computation. And with each new language came more powerful levels of abstraction, such as object-oriented programming, which is now one of the most widely used programming constructs. In addition, these newer models of computation often are developed to solve common problems. Some of these newer models of computation are developed for general- purpose programming tasks, but some are developed for a particular application. For example, LabVIEW was developed for test, measurement, and control applications, and Python was developed for quick code-scripting tasks. These increasing levels of abstraction result in languages better suited to specific tasks. The best test managers now design test systems that leverage the power of multiple languages and save development time by using test management software. Traditional Test System Development Considering that software is the backbone of automation when building a test system, many organizations prefer to standardize on a single, fairly general language used for all aspects of test system design from individual component tests to test management across the board. The end result is the development of a homogenous test software approach. The main advantage is all members of a team can work in a single, standardized environment, which allows easier sharing of libraries and code modules across the team. The training for this approach is also greatly simplified since the team learns and works in a single environment. However, standardizing on one language does present some disadvantages. Using a single language can limit new hires to a certain skillset or force new employees into learning new tools. This topic of building a skilled workplace was explored in NI’s Automated Test Outlook 2014. As students graduate, they often have a preference for and experience in one or more specific languages. Also, as new managers take over, they commonly opt to implement a language of their choice, which causes The Rise of Test Management Software Look around your test group. If you are like most organizations that have conducted automated test for a while, you’re probably seeing an increase in languages. Thanks to more specialized forms of abstraction in today’s higher level programming languages, this problem is not going away anytime soon. It is only intensifying. “We approach unique test challenges with the best language for each task at hand. By using commercial off-the-shelf test management software to act as a grand unifier during the whole product life cycle, we greatly increase our engineering productivity and minimize the time to market.” — Simon Wiedemer, Head of Automated Testing Architecture, Festo AG Co. KG ni.com/ato 09 AUTOMATED TEST OUTLOOK 2016
- 7. ACCELERATED PRODUCT DEVELOPMENT THROUGH STANDARDIZATION Time to Market Correlate Results Test DUT Choose Hardware Write Software Characterization Verification and Validation Preproduction High-Volume Manufacturing Characterization Verification and Validation Preproduction High-Volume Manufacturing STANDARDIZEDTRADITIONAL Two decades later, a quad-band “world phone” costs a few hundred dollars. Even a basic mobile phone that supports over 20 cellular bands, in addition to Bluetooth, Wireless LAN, and GPS technology, retails for under $100 today. While this dramatic drop in price has significantly benefited consumers, it has created substantial challenges for those supplying the RF components inside. According to a recent Databeans analyst forecast, the cost of radio frequency integrated circuits (RFICs) for mobile devices has dropped by more than 40 percent since 2007. This price decrease is coupled with the challenge of rising device complexity. Ten years ago, a single-function GSM power amplifier was the norm. Today, many RFICs are significantly more complex. They support multiple radio standards and multiple bands with more advanced technologies such as dynamic power supplies, MIPI digital interfaces, and more. To maintain margins while the average sales price shrinks, companies must reduce the cost of semiconductor design and test. Given that test cost is often nearly half of the cost of goods sold, according to IC Insights, RFIC suppliers have a renewed focus on decreasing the cost of manufacturing test. Over the past decade, this intense focus has produced a significant shift from using turnkey ATE solutions to building in-house and cost-optimized testers based on off-the-shelf instrumentation. The ability to specify a tester for a specific IC device—along with improvements in instrumentation technology—can significantly reduce test cost. This shift to a custom tester approach has been a large factor in the success of modular instrumentation platforms like PXI in manufacturing particularly because modular instruments have shown excellent value per performance. Competition and Innovation Increase Pressure on Cost As intense market competition and an ever-accelerating pace of wireless innovation complicate the desire for lower costs, recognizing that these forces require shorter product release cycles for companies to stay competitive is important. With much of their manufacturing test costs already reduced through the use of modular instruments, organizations must find new ways to improve the efficiency of product development. Once considered the holy grail of product development, one increasingly important practice is to shorten the product design cycle with standardized design and test tools. In the past, product development teams often used different design and test practices and equipment in each phase of product development. With this approach, the engineer validating silicon and the engineer designing the manufacturing test plan were often left to design their own tester from the ground up. To be profitable, companies must improve efficiency between each phase of the product development cycle. As a result, many companies are adopting integrated Standardizing Platforms From Characterization to Production In 1983, the first commercial mobile phone retailed for $3995, almost $10,000 in today’s economy. It supported a single band, weighed almost a kilogram, and was about the size of a brick. ni.com/ato 11 platform approaches to help reduce total test costs as well as shorten time to market. As the 2015 McClean Report says, organizations must place greater effort into “decreasing IC design, development, and fabrication expenditures in order for the industry to maintain its continuous reduction in cost per function.” Although the desire to use a common test platform from design to test is not new, innovations in test equipment are now making that possible. A decade ago, the test equipment engineers might have used in their characterization labs was simply not fast enough for high-volume manufacturing test, and using different tools throughout the product life cycle was a necessity. Today, PXI instruments offer the measurement accuracy required for RD and the speed required for manufacturing test. As a result, organizations are increasingly standardizing on modular instrument platforms throughout the entire design cycle, which directly reduces the cost associated with correlating measurement results. In addition to the improved speed and measurement quality of PXI, application-specific systems, such as the NI Semiconductor Test System, build on the PXI platform by adding a rugged enclosure, fixturing, DUT control, and the turnkey software required for the semiconductor manufacturing environment. Mergers and Acquisitions Drive Standardization Other factors driving the need to standardize on common design and test platforms are mergers and acquisitions within the semiconductor industry. Although the consolidation of suppliers enables companies to address a larger set of components in a particular mobile device, it uniquely impacts the engineering teams responsible for delivering products to market. This impact is usually caused by merging geographically distributed engineering teams that have their own preferences in programming languages, test strategy, and tool investments. Product development inefficiency often emerges when distributed teams do not share common best practices. As a result, many organizations are in the midst of standardization. One critical focus is using a single codebase from automated measurements in RD to automated measurements in manufacturing test. By sharing a common codebase of test software, along with using the same core measurement technology throughout the design cycle, organizations have reduced test software development cost and ultimately decreased time to market. Status Quo Leaves Money on the Table Just as the digital age commoditized digital IC, the information age is commoditizing analog IC. Commoditization comes with lower cost, and it requires a dramatically new approach to test. In an era where test strategy is considered a competitive advantage, organizations are using standardization on a common platform as a method to reduce test costs. However, if an organization is not considering a common platform approach for test, it might be in trouble. Although the old way is sometimes easier, the additional cost and inefficiency leave company profits on the table. AUTOMATED TEST OUTLOOK 2016 “…greater effort is being placed on decreasing IC design, development, and fabrication expenditures in order for the industry to maintain its continuous reduction in the cost per function.” —The McClean Report 2015, IC Insights
- 8. In addition, the proposed study of frequencies ranging from 24 GHz to 86 GHz from World Radiocommunication Conference 2015 shows the importance of mmWave frequencies for use in 5G wireless communications. The benefits of extending communications networks to mmWave are derived from the physical characteristics of operating at such high frequencies. These include greater spectrum efficiency from wider bandwidth signals, shorter wavelengths that require smaller component dimensions and allow more antennas to be packed into the same space, and the ability to handle larger network capacity and better isolation of simultaneous users. Consumers reap material benefits but RFIC manufacturers face new challenges when they cannot directly transfer the methodologies used to test previous standards. Leading RFIC manufacturers will have to adopt modular hardware and scalable software to test mmWave frequencies successfully both financially and technically. The Impact on Business Strategy The substantial benefits of mmWave technology in consumer applications point to a complex and economically challenging portfolio of RFIC designs necessary to serve a variety of niche applications. RFIC manufacturers are now thinking about economies of scope as well as economies of scale. Semiconductor manufacturers must either institutionalize the reuse of common functional elements across multiple designs or accept lower margins to remain competitive. Economic considerations are further magnified when designing for an emerging standard like 5G. Divergent frequency band proposals from multiple parties generate uncertainty as to which spectrum will ultimately be included in the standard. Unless test systems are flexible, investing before standards are finalized creates a high risk of sunk costs. To mitigate these high initial costs, manufacturers have used test methods like “golden DUT” and built-in self- test. But these low-cost options aren’t enough to fully validate designs. True metrological rigor must be applied to ensure that products can be certified and can conform to specifications and regulations. This can be achieved only through performance validation using traceable test instrumentation. Maintaining test flexibility is difficult because of the costs of supporting new frequencies and bandwidths as well as the costs of covering investments made for frequencies that are ultimately orphaned. Similar to the functional reuse underlying successful IC design strategies, investing in the right test system means investing in modularity. An economically efficient test system provides a core platform by isolating functional components for modulation, demodulation, data movement, and processing. This platform can be paired with the appropriate extension to reach the frequency of choice. Such a design allows investing in components that scale across frequencies. It also helps mitigate the costs associated with serving various applications and Making (mm)Waves in Test Strategy Expectations are high for the integration of millimeter wave (mmWave) technology into the connectivity and communications protocols defined by the IEEE and 3GPP. WiGig (802.11ad) is already delivering extremely high-throughput and low-latency connectivity for numerous applications. “Millimeter wave research has presented a myriad of technology proposals, which continue to evolve, and since frequencies of operation and bandwidths have not been finalized, flexible systems for characterization, VV, and production testing will prove to be invaluable in their ability to remain nimble during the establishment of standards.” —Amarpal (Paul) Khanna, PhD, IEEE Fellow, General Chair for IMS 2016, and RD Manager at NI MODULAR SYSTEMS EASILY SCALE FOR mmWAVE Sub 6 GHz Sub 6 GHz and mmWave ni.com/ato 13 Trend Watch 2016 accounting for the uncertainty in 5G development while offering fully conformant test capability. Test Challenges: The Lab Versus the Floor Unlike sub 6 GHz frequencies, antenna dimensions at mmWave frequencies are quite small. They are often fabricated as patch arrays bonded to the IC itself, which makes transmission technologies like beamforming more economical but challenging to characterize. As is common in design validation, devices are placed in anechoic chambers with an array of test antennas in locations that validate emissions patterns over the air (OTA). Though OTA test systems are essential for the complete validation of device performance, difficulty in ensuring radiated isolation, the need for high-touch test procedures (reduced throughput, increased cost), and the physical cost and size may make them unsuitable for the production floor. The difficulty in transferring OTA-based test systems from RD into production makes clear that effective test solutions must take advantage of core functional hardware and software platforms and provide flexible connectivity to the device. Using common core platforms with different connectivity, manufacturers can run statistical correlations on the results collected from high-sample-count OTA tests in the lab. These results validate radiation patterns under different beam configurations, with probed results produced by individual antennas during the same beam configurations. By using the common back end, manufacturers can deploy the appropriate front-end solution for each test scenario without having to invest in a new test system configuration for each stage of the value chain. Multifunction Designs Yield New Test Considerations Another disruption to old test methods is the integration of mmWave technologies with other standards to form a “multiband” wireless system that schedules users and data more efficiently. Good examples are wireless routers touting “triband” Wi-Fi, though today they actually use one and two channels in the 2.4 GHz and 5 GHz unlicensed band, respectively. With the addition of 802.11ad at 60 GHz, this system will truly be “triband.” Testing such functionally integrated devices poses new questions. Do these become single modules that support all bands in one package? What test points will be exposed for functional test? Will these modules be tested at all bands on one station, or will they separate sub 6 GHz test and 60 GHz test? Modular platforms like PXI mitigate these scalability challenges by simplifying new I/O insertion that can augment existing functionality. In addition, traditional single-standard systems, which commonly incorporate LAN buses, do not scale to handle larger waveform sizes limited by data movement and processing power. For example, a typical 802.11ad signal requires up to 50X the data movement capacity as a typical 802.11ac signal. Unlike traditional systems that quickly grow in size and cost, modular test platforms like PXI provide a well- balanced combination of size, cost, and I/O to support functionally integrated applications. PXI can handle nearly 16 GB/s of throughput today, and, as the underlying bus and processing technologies continue to evolve, it seems to be the logical test platform for future test technologies. Test Strategy = Business Strategy With the heavy investment in mmWave research for 5G communications and the remaining uncertainty around the frequencies that will emerge to support this rollout, it’s clear that the challenges WiGig RFIC manufacturers face today will only grow in scope. Savvy organizations will plan test strategies early through close partnerships with vendors that can offer flexible test platforms supporting long-term product strategies. Modular platforms have led the research of 5G and mmWave technologies and will continue to lead the charge in test by providing cost-effective, technology- flexible solutions for the growing complexity of applications using high-frequency technologies. AUTOMATED TEST OUTLOOK 2016
- 9. US Corporate Headquarters 11500 N Mopac Expwy, Austin, TX 78759-3504 T: 512 683 0100 F: 512 683 9300 info@ni.com ni.com/global—International Branch Offices ni.com/ato ©2016 National Instruments. All rights reserved. Big Analog Data, LabVIEW, National Instruments, NI, and ni.com are trademarks of National Instruments. Other product and company names listed are trademarks or trade names of their respective companies. 351409F-01 24244