Pages Menu
TwitterRssFacebook
Categories Menu

Posted in Uncategorized

August Q&A Interview

Jin Yokoyama, Functional Manager of Memory Test, Advantest
By GO SEMI & Beyond staff

Advantest recently announced the T5822, the latest addition to its highly successful T5800 series of memory testers. In this issue’s Q&A, Jin Yokoyama, Functional Manager of Memory Test, talks about the new tester, which performs wafer-level test of DRAMs, NAND flash devices, and other non-volatile memories (NVMs).

What market demands prompted the development of the T5822?
Semiconductor makers need low-cost solutions for high-volume testing of price-sensitive memory ICs. The primary driver is the booming mobile-device market. While the PC market’s share of DRAM bit demand plunged to just 26 percent between 2009 and 2016, mobile DRAM bit share has grown more than 500 percent since 2009. Add in the fast-growing multilayer 3D NAND market, and you have a clear need by manufacturers of multiple memory devices for a cost-effective test system that can handle their requirements with a range of capabilities.

How does the T5822 fit with other current memory testers in the T5800 family?
The T5822 is designed for economical wafer-level testing by IDMs/OSATs [integrated device manufacturers/outsourced semiconductor assembly and test providers] producing or testing both DRAMs and NAND flash memories. When it comes to test coverage and functionality in between the T5822 and our existing T5833 in the wafer-level testing segment, some features do overlap. However, they can viably co-exist in the line because each system’s strength is different. For example, in addition to wafer sort, the T5833 is capable of higher-speed interface testing at the package level (i.e., final test) – it’s essentially a superset of capabilities. But at the wafer level, the T5822 will provide a wider voltage swing for some resources, allowing it to cover a wider range of customer needs – it provides greater optimization for memory wafer testing.

What degree of parallel test does the new system offer?
It depends upon each device interface specification or test mode. In general, the T5822 will be able to support up to 1,536 devices under test (DUTs) in parallel per device, and at high speed – up to 1.2 Gigabits per second.

In what geographical region do you expect demand for the T5822 be greatest?
Judging by recent M&A activity in the memory market, the pool of key memory makers/OSATs is growing more oligopolistic. Thus, we can’t say too much in this regard, as it would likely be obvious what companies we’re referencing. Generally speaking, we are expecting interest from customers that are dealing with both DRAM and NAND memory manufacturing (wafer-level testing), such as combo IDM/mixed-memory-product OSATs in the U.S. and Korea.

What are the three most important points you’d like readers to know about the new product?

  • The T5822 has been optimized with a wider voltage level swing so that the T5800 series can fully cover all memory wafer-testing segments.
  • The system has been enhanced with a compact test head that is more economical compared to our legacy general-purpose solution in our other T5800 series testers. It can be configured with a single or dual satellite test head, the latter being the most common configuration.
  • The tester is part of the proven T5800 scalable and flexible platform family, which provide high reliability and enable easy test program porting from other legacy platforms due to high software compatibility. Together with modularity, FutureSuite operating system, and memory redundancy analysis (MRA) software, these are key advantages for our target customers.
Read More

Posted in Featured Products

Five Key Success Factors for Automotive Device Test

By Kotaro Hasegawa, Senior Director, ADS System Planning Department, Advantest Corp.

 

The number of automotive-related semiconductor devices being designed into vehicles has been growing rapidly due to increased requirements associated with safety, in-vehicle entertainment, and advanced driver assistance systems (ADAS; see Figure 1). As ADAS content advances, we move closer to autonomous vehicle driving (Figure 2), which will require even more devices.

Figure 1. Automotive-related device market trends (Fuji Chimera Research Institute, Inc., 2017)

In turn, as this demand, goes up, production volumes for application-specific standard products (ASSPs) need to escalate. Thus, more cost-effective testing solutions are required to improve cost of test (COT) while maintaining test quality. Automotive devices must work all the time – this is a critical safety factor, made even more so in the ADAS era. Thus, maintaining high device quality in mass production is essential to developing automotive ASSP device test cells.

Below are five key factors associated with achieving testing success and helping ensure high quality as emerging automotive ASSPs enter into mass production.

Figure 2. Definition of ADAS levels (copyright SAE International, 2014)

1. Optimized Parallelism Optimized Parallelism
Typically, when testing system-on-chip (SoCs) on automated test equipment, parallelism is mainly defined by resource counts. To increase the amount of parallelism, you simply add further resources onto the tester and align them with the prober or handler. Automotive ASSPs, however, involve a large amount of analog testing using an external control circuit due to the need to simulate an application model for high-quality testing. Performance board (loadboard) space is required to mount all components necessary for testing the device under test (DUT).

In many cases, even with sufficient resources on the test head, loadboard space limitations prevent achieving the desired level of parallelism, which, in turn, limits productivity. One way to address this challenge is to expand the printed circuit board (PCB) area to ensure enough space is available to mount all components, but sometimes even this step is not sufficient.

The Advantest T2000 test platform incorporates two capabilities that address these parallelism limitations, allowing it to achieve greater parallelism than competitive platforms:

  • Reduction in application relays on loadboard – up to 40 percent fewer in some cases – provided by the T2000’s multifunctional modules (Figure 3).
  • Wider PCB support with RECT550EX fixture (HIFIX) – this feature can enable a 140-pin automotive ASSP device to expand x4 DUT testing to x8 DUT testing, as an example (see Figure 4).

Figure 3. Example of T2000 relay reduction multifunctional module

 

Figure 4. T2000 RECT550EX fixture (HIFIX)

2. Reduced Test Time Reduced Test Time
Once devices enter high-volume manufacture, test-time reduction becomes even more critical. The T2000 includes some unique features to improve productivity while maintaining test quality, enabling best-in-class performance.

Typically, test systems require the user to power off, change mode, and power on again, but this takes a great deal of time. The T2000 architecture’s switching methodology reduces power spikes by enabling mode switching during test. This reduced switching time is ideal for devices that are extremely COT and time-to-market (TTM) sensitive, such as automotive ASSPs.
Another testing method is to monitor the output behavior of the DUT. With the T2000, the hardware module can automatically detect the device output, and if there is any change in device state, the module can halt the arbitrary waveform generator’s (AWG) input as needed and assess the test result, then immediately move on to the next test item (see Figure 5). This capability speeds the process because waiting to receive the full-scale AWG output isn’t necessary.

Figure 5. Test time reduction method on AWG

3. Wide Coverage Wide Coverage
The T2000 features 52 test head slots, allowing a reach of more than 8,000 digital and 6,400 analog channels. This lets the user obtain the best resourcing fit, choosing from wide variety of T2000 modules available. In addition, the T2000 can deliver both high voltage (up to 2000 V) and current (up to 216 A). With this wider coverage, the platform can be utilized to test automotive ASSPs, power management integrated circuits (PMICs), light-emitting diode (LED) drivers, and other devices with high resource requirements (see Figure 6).

Figure 6. The T2000’s test segment coverage is among the industry’s widest

1. Test quality
Advantest integrated into the T2000 platform a long lifecycle, as well as testing stability, accuracy and reliability. The platform and modules possess unique hardware/software design rules in order to achieve high-quality hardware. All modules are developed based on these design rules, including selection of the ICs for the module, to help ensure high system quality and reliability. These are key factors for testing quality-sensitive devices such as automotive ASSPs.

5. Field Proven
It takes a long time to launch new automotive-related devices into the market, and the test process is highly sensitive to changes in the production environment. Selecting a proven platform that includes all necessary capabilities is key to optimal testing of automotive-related devices. In addition, newer devices targeting this market will be highly integrated to include more functionality, voltage coverage, current, and frequency. In addition to being field-proven, the test system implemented will need to offer the widest coverage so that it is flexible and extendible.

Worldwide, more than 3,900 T2000 modules designed especially for automotive ASSP testing are already installed, and further growth is expected as the market for these devices continues to expand.

Given its combination of high-quality test, high productivity and widest device segment coverage, the T2000 is well suited for testing automotive and other quality-sensitive devices.

 

Read More

Posted in Upcoming Events

Advantest to Debut New Solid-State-Drive Testers and Present Technical Papers at Flash Memory Summit, August 8-10 in Santa Clara, California

 

Advantest Corporation will showcase the latest additions to its MPT3000 series of solid-state drive (SSD) test solutions and present two technical papers at this year’s Flash Memory Summit on August 8-10 at the Santa Clara Convention Center. Advantest is an emerald sponsor of the 2017 Summit.

In booths #606-608, Advantest will present two new options for its MPT3000 platform, adding to the range of SSD protocols and form factors that it supports. The new MPT3000HES leverages the same hardware, software, thermal performance and interfaces used throughout the MPT3000 series, but in a smaller configuration designed for “copy exactly” engineering and lower-volume applications. In addition, the new Smart Power option is a highly cost-effective solution for built-in self-test (BIST) needs, including production test insertions in which a low-speed serial interface can provide maximum synergy with protocol test insertions.

Advantest’s exhibit will also feature a digital display and laptop demonstration of the cost-efficient T5851 tester for universal flash storage (UFS) devices and PCIe BGA memories. This system uses the same proven test architecture at the MPT3000 product family.

The company’s technical experts will make two presentations at the summit during Session 201-C on Testing Issues, which begins at 8:30 a.m. on Wednesday, August 9. Advantest’s Vishal Devadiya will address “Preparing SSDs for Qualification Testing” and Ben Rogel Favila will discuss “A Scalable Platform for Optimal SSD Test.”

Read More

Posted in Upcoming Events

Advantest Shines at Semicon West 2017

The annual SEMICON West show, held July 11-13, 2017, at the Moscone Center in San Francisco, proved once again to be a great venue for interacting with customers, press and analysts, as well as promoting Advantest’s extensive product portfolio.

This year, Advantest showcased its new T5822 memory tester and V93000 Wave Scale™ MX-HR card, as well as its system-level test (SLT) application presentation. The Wave Scale RF and MX-HR modules and T2000 application module on-board (AMO) components were shown in the booth, using organic light-emitting diode (OLED) screens to highlight promotional messages interacting with the products behind the transparent glass. Also featured in the booth were the EVA100 mini shoebox configuration, corporate presentation, and Wave Scale videos.

In addition, Advantest participated in SEMI’s new SMART Journey Pavilion, which allowed visitors to explore the Internet of Things (IoT) and other “smart” innovations that are revolutionizing the manufacturing supply chain, automotive applications and everyday life. In this pavilion, Advantest screened a dynamic video illustrating the numerous devices that we test in autonomous vehicles.

Advantest presented four technical papers during the TestVision 2020 workshop, held July 12-13 in conjunction with SEMICON West. Presenters were: Dave Armstrong, director, business development; Roger McAleenan, director, Millimeter-Wave Test Solutions; Adrian Kwan, manager, business development; and Kotaro Hasegawa, senior director, ADS System Planning Dept.

Our annual SEMICON West customer event was again held at the popular and vibrant 111 Minna Gallery. More than 200 attendees enjoyed an evening of socializing in an informal setting, with entertainment provided by San Francisco’s own celebrated contemporary violinist Gabi Holzwarth.

Next year, Advantest will be located for the first time in the brand new South Hall at Moscone, which is currently under construction. We look forward to sharing with you our latest developments and innovations at SEMICON West 2018!

Gabi Holzwarth performing at 111 Minna Gallery during Advantest’s Customer Hospitality Event

Read More

Posted in Top Stories

5G Lessons Learned from Automotive Radar Test

By Roger McAleenan, Director, Millimeter-Wave Test Solutions, Advantest America

Situated between microwave and infrared waves, the millimeter-wave spectrum is the band of spectrum between 30 gigahertz (GHz) and 300GHz. It is used for high-speed wireless communications and is widely considered as the means to bring 5G into the future by allocating more bandwidth to deliver faster, higher-quality video, and multimedia content and services. Automotive radar is the entry point into millimeter wave for testing purposes.

Automotive radar has been evolving for the past several years, with Tier One companies producing and developing designs for a variety of different applications. As automotive is considered one of the key vertical markets for 5G technology – others include mobile broadband, healthcare wearables, augmented and virtual reality (AR/VR), and smart homes – radar systems in vehicles can provide valuable insight into the other millimeter-wave applications.

The 5G standard promises new levels of speed and capacity for mobile and wireless communications with greatly improved flexibility and latency compared to 3G and 4G/LTE technologies. However, its unique chip structures will create new challenges for test and measurement. By understanding the limits of test equipment, systems and hardware, we can better address the practical aspects associated with delivering on the promise of this technology.

Test and measurement challenges
From a measurement perspective, 5G and auto radar have functional characteristics in common that need to be measured, such as signal blockage, radiation interference and beamwidth selection. Another aspect is loss of signal penetration, an area where radar has an advantage over optical techniques that can be confounded by rain or snow. The band assigned to automotive radar, 76-81GHz provides greater accuracy in range resolution, and is sandwiched between point-to-point (P2P) bands on each side.

The challenges to be addressed in 5G test are similar to those associated with automotive radar, as well. Challenges in millimeter-wave applications include:

  • Handling multiple port devices economically
  • Providing features and testing optimized for characterization and production
  • Over-the-air environment due to packages with integrated antennas
  • High-port-count switching/multiplexing (4×4, 8×8, etc.), often in the same device
  • High levels of device features on a die– MCU + memory  + radio + high-speed digital

Multiple antennas improve power efficiency since more energy is pointed where it needs to be, and with steering, multiple targets can be tracked. This provides improvements to the capabilities and applications expand broadly to “surround” safety features, vehicle-vehicle coordination/communication.  The increased complexity in devices extends up to multiple combinations of transmit and receive.  This functionality will significantly improve vehicle-to-animal/human/object recognition and avoidance, as well as tracking more targets simultaneously.

Transceiver design is important, and they can be optimized as required as a low- or zero-intermediate frequency (IF) design. Automotive and 5G radios look nearly the same, with the similar IP blocks, e.g., phase shifters, local oscillators, RF amplifiers and mixers (Figure 1). The primary distinction is 5G radios’ modulation capability. Both may include up and down conversions, but for 5G, the market is looking for information bandwidth increase. This is actually pretty difficult from a test perspective because it requires elaborate analog equipment like high-performance oscilloscopes. This aspect is still a work in progress.

Figure 1. Transceiver design in automotive and 5G systems is highly similar.

Four main millimeter issues and considerations must be addressed in auto radar. This applies to 5G as well in that these four problems – rain attenuation, Fresnel zone, path loss, and ground reflection – are all problematic, whether you’re driving a car or the equipment is on a tower. Figure 2 shows all the areas in which radar is being used in cars, and further underscores the challenges associated with effective testing of these systems from a system level perspective.

Figure 2. Radar zones in vehicles continue to multiply as automated content increases.

One way to address some of the operational millimeter challenges is through beamforming. This is a technique that focuses the radar transmitter and receiver in a particular direction. Beamforming can be passive or active, although the former is limited in its effectiveness. Active RF beamforming, the increasingly preferred approach, will be gamechanging: it enables tracking multiple objects, both moving and static (people, vehicles, buildings, etc.) at various speeds, simultaneously. This allows auto radar to actively steer the beam toward objects and track them independently. Because the beam can be positioned with so many possibilities, testing in this way is currently a rhetorical question, although several automakers are working on solutions. For 5G, the beams would normally point either to other towers or to individual handsets and be able to track them. Basestations will have antenna arrays that can be steered to track people with 5G handsets – this will be an essential success factor in achieving the information bandwidth promise.

Test lessons learned
Advantest’s automated test equipment has been deployed for testing automotive radar for more than four years, testing from 18GHz to 81GHz, including wireless gigabit (WiGig) test in the 60GHz range, which may also be applicable to 5G.

At the moment, the focus remains on device test, but this is changing. Millimeter-wave applications provide an ideal opportunity to move away from component-level test and more toward higher-level models and end-to-end system-level testing. Figure 3 highlights the growing trends associated with system-level test. With that noted, here are some key lessons learned from Advantest’s work in the auto radar space, using its proven V93000 test platform.

Figure 3. Demand and opportunity for system-level testing is on the rise.

  1. Power accuracy is critical. This will be very important to understand and address because, as we move closer to built-in self test (BIST), the device must be able to measure accurately the power it’s generating. Right now, we’re still learning how to get RF CMOS and BIST working together to give an accurate power measurement.
  2. Metrology is difficult. Given the various connectors and waveguides that must be navigated, there are few reliable ways to perform accurate metrology of fixtures, connectors, loadboards, and other components. Also, there is the issue of system degradation – every time a new part is tested, it degrades slightly due to the materials used, and over time, the sockets or membranes that begin to deteriorate. In addition, when something finally needs to be changed out on the test system, recalibration must be performed, and that can cause a slight change in measurement results when combined with the degradation issue.
  3. Limits need to be established. As devices grow more complex and better – and as efforts are made to extend radar range – two key factors come into play:
    • Phase noise – This key parameter on RF signals affects performance of radio systems in various ways. It’s important to understand at what point phase noise begins to impact performance    and the cost-benefit.
    • Noise figure – This measure of the degradation of the signal-to-noise ratio, caused by components in an RF signal chain, is essential to making radar more effective. The key question in this regard is, what’s the smallest signal I can see (relates to dynamic range)?
  4. Millimeter “anything” is expensive. Currently, there are significant costs associated with millimeter-wave technology that will likely decline over the next few years. In the meantime, some chipmakers are trying to implement millimeter-wave technology for smaller end products, such as radar distance measuring devices, but they can’t build them because they can’t figure out how to test them economically on a small scale. The solution may rely on future technology that is still being developed.
  5. Test engineering knowledge is scarce. This is perhaps the most critical factor of all – hence, saving it for last. The number of engineers working in millimeter technology is relatively small, and companies wanting to enter the space can’t simply materialize engineers versed in radar technology to help them with product development – particularly when the primary emphasis in most engineering programs is digital technology, rather than analog/RF. This means that talent is expensive, which can put a real damper on what companies are able to do. We need competent engineers to be trained that are strongly motivated and passionate about millimeter-wave.

Summary
Automotive radar technology is here now, and while it’s currently being seen primarily in premium-brand vehicles, the goal is to bring down the unit cost so that it becomes standard equipment throughout the automotive industry. To do this, a number of challenges must be addressed, including solving of the complexities associated with testing. Advantest is strongly committed to this market and in taking a leading role in finding these solutions and applying them to other millimeter-wave applications as the market continues to grow – including the fast-emerging 5G.

Read More