Mar 06, 2007This article was originally published by RFID Update.
March 6, 2007—Intel, arguably the world's most recognized chip company, today announced the release of a fully integrated Gen2 RFID reader chip, the R1000. The new chip, which is slated for volume production within weeks, may represent a major technological milestone for RFID as it precipitates a dramatic reduction in the cost and form factor of reader technology. RFID Update spoke with Kerry Krause, Intel's marketing director for RFID, about the R1000 and his company's expectations for its impact.
With few exceptions, RFID readers today are comprised of multiple electronic components from different manufacturers. These components are purchased by the reader manufacturer, then assembled into a cohesive whole to create the final reader product. What Intel has done in the R1000 is design a single chip that provides most of the same functionality, but cheaper and smaller. "If you crack open any of the leading readers," Intel's Krause explains, "you see a very complex and expensive design. The R1000 integrates about 90 percent of all those discrete components onto a single chip."
The cost and size reduction of the R1000 could have dramatic consequences for RFID readers, according to Krause. For example, the standard fixed Gen2 readers commonly deployed in distribution centers and retail back rooms currently sell for roughly $1,000, depending on volume. Krause expects that the savings provided by the R1000 could cut that cost in half in the next ten months. "I expect to see that those high performance Gen2 UHF readers will be $500 or below by the end of the year."
Furthermore, the smaller form factor will allow the installation of readers in more places, potentially breaking open the ways in which RFID is applied. "You're going to see a lot of innovative readers in more focused spaces and unique applications," asserts Krause. "When products like those hit the market, it's going to be interesting to see what people do with them."
Yet another benefit that Intel foresees is improved compatibility of core reader functionality across vendors. As RFID reader companies base their products on the R1000, the core reader technology and functionality will be the same, not unlike how Dell, HP, IBM, and Sony computers use the same Pentium-based chip technology. This sort of competitive landscape provides a level playing field for RFID reader vendors to focus on added value -- rather than low-level reader processor design -- for competitive differentiation. "The R1000 offers the common building blocks," explains Krause. "It lets reader companies focus their R&D efforts at a higher level of the stack, and it provides cross-vendor compatibility that tends to be good for market growth and technology adoption."
Intel has already been working actively with many reader companies to incorporate the R1000 into iterative models or even entirely new reader products. "We have been enabling our customers [reader vendors] for about eight months," indicates Krause. "Twenty customers are developing product for the R1000 today, ten of which will have prototypes on display at RFID World in Dallas [at the end of this month]." A number will have commercially available readers in Q2, including Alien, AWID, and ThingMagic in the US, CAEN and Deister in Europe, and AMOS, Chung Nam Electronics, Kenetics, MTI, Samsung-Techwin, and Unitech in Asia-Pacific. Krause noted that beyond those companies, some of the other recognized RFID reader companies will "fast follow" with R1000-based products in the months ahead.
In addition to hardware companies, Intel has been working behind the scenes with leading enterprise software companies to add support for the R1000's application programming interface (API), which allows packages from the likes of Microsoft, SAP, Oracle, IBM, BEA Systems, Reva Systems, and noFilis to communicate with the R1000.
"Right at launch," says Krause, "we're going to have ten partners with products, and broad support from the software community."
Intel is targeting the R1000 at all flavors of Gen2 RFID readers, from fixed to handheld to embedded. This is in contrast to the Gen2 reader chips that have already been announced by competing vendors like WJ Communications, Starport Systems (announced just yesterday), and Singapore's Institute of Microelectronics (see Compact, Cheap RFID Reader Chip Developed). Because of power limitations, those chips are primarily targeted at handheld readers. (Handhelds are designed for use at close range and therefore require less power than their fixed counterparts, which are designed to have read ranges of many meters.)
Krause noted that despite the industry's focus on supply chain- and retail-oriented fixed readers, Intel believes that demand for other types of readers will surpass that of fixed readers in the not-too-distant future. "For the industry overall, you're going to see that embedded modules, handhelds, forklifts, and other reader variants will eventually greatly outnumber dock doors," he predicts. "You're going to see a lot of innovation in the reader space. Outside of the supply chain, they're pursuing a wide range of really innovative end user applications for RFID."
The R1000 was incubated within Intel Capital. "RFID started as a little seed business over three years ago," explains Krause. "Intel identified RFID as a compelling technology that could have significant impact." Once that impact was determined, the company began assessing whether a product opportunity existed that was suited to Intel's core competency in complex chip manufacturing. Tag chips (like those produced by Impinj, Philips, STMicroelectronics, Texas Instruments, and others) were considered but ultimately decided against because of their relative simplicity. "While the volumes are enormous for tag chips, it tends to be a more simple design, which is not necessarily a good fit for our technical expertise nor for our business model."
Reader chips, on the other hand, were. Development of the R1000, from when Intel engineers first started architecting it until its volume production set for the end of this month or the first week in April, will have taken two years. Krause would not comment on how much capital has been invested in the venture, saying only, "It's not insignificant. It's what would be typical for the development of a complex semiconductor product."
Given the investment of time and resources, it stands to reason that Intel could be concerned that the RFID market's real growth curve has been far more gradual than projected when chip development began two years ago. "The projections have changed," acknowledges Krause, "but we're still bullish. Right now the trend is looking good in terms of the quality and quantity of customers that are supporting the R1000, and the aggressiveness with which they're developing product."
Asked whether the R1000 represents the kind of leap in reader technology that will accelerate RFID adoption, Krause answered, "I think so, and I sure hope so."