OCTOBER 16, 2007 By Stephen Hardy -- Looking to plug a hole in its 40-Gbit/sec offerings, JDSU (search for JDSU) has partnered with 40-Gbit/sec pioneer Mintera (search for Mintera) to co-develop products, starting with a 300-pin transponder module based on the latter's adaptive differential phase-shift keying (ADPSK) modulation technology. The partnership includes a JDSU investment in Mintera.
The amount of the investment was not disclosed. (See story on Mintera's funding announcement.)
According to Craig Iwata, director of product marketing at JDSU, the partnership paves the way toward an integrated product offering for 40 Gbits/sec. This includes ROADM and EDFA transport elements as well as test equipment in addition to the upcoming module, which should be available during the first half of next year.
The partnership will benefit both companies, Iwata believes, in terms of positioning and future product development. "We'll essentially be leveraging Mintera's expertise in the transmission space, JDSU's expertise in the transport space, as well as our manufacturing and supply chain leadership," he says. This last aspect addresses concerns some systems houses may have about turning to a relatively small, young company such as Mintera -- and, by implication, Mintera's chief rival, StrataLight Communications (search for StrataLight) -- for its 40-Gbit/sec technology requirements.
Iwata adds that, in his opinion, JDSU's new combination of capabilities puts them in the best position to serve the needs of network equipment manufacturers and the carriers they serve. "When people make transport decisions they have to think about their transmission. And vice versa -- when people make transmission decisions they need to keep in mind what their transport looks like," he says. This thought also will apply to the two new partners, who expect their collaboration will result in improved products for both companies.
While the partnership will focus on the transponder initially, the two companies also expect to co-develop additional products. Iwata did not speculate on what those products might be.
The upcoming ADPSK transponder will be based on a design already under development at Mintera, says Niall Robinson, Mintera's vice president of product marketing. Robinson says the module had already garnered "major" customer design wins. He declined to say whether those wins had come before or after news of the partnership had been shared with potential customers. "I can say is that they are very happy with the partnership," he granted.
Mintera introduced ADPSK technology to the market in February with the release of its MI 40000XS platform; the company also has made custom boards leveraging ADPSK that would fit into the chassis of other system houses. ADPSK is an enhanced version of the company's proprietary variant of DPSK, which they call partial DPSK. (See our story that describes partial DPSK.) DPSK does a good job of providing extended reach in comparison to such "first generation" 40-Gbit/sec modulation formats as optical duobinary and carrier-suppressed RZ. However, it runs into problems at line widths narrower than 100 GHz. Both Mintera and StrataLight (among others) have experimented with variants of DPSK to overcome this drawback. Partial DPSK provided good performance, Robinson says, but required that the transmission element be set according to the number of filtered elements (such as those within ROADMs) the signal would encounter in the link. This necessitated multiple part numbers and an inventory headache. ADPSK, as its name implies, can adapt to a variety of element combinations.
Robinson says the module will be made available initially in an NRZ format, which he believes the industry currently would prefer.
The transponder will not contain any dispersion compensation technology, although dispersion compensation will be a requirement. "You'll talk to N number of customers and you might get N results in terms of their preferences for the dispersion compensation technology," Robinson explains. "I don't think the industry has quite settled on which dispersion compensation technology is going to win out long term. So, for now, it's actually almost a market requirement that the dispersion compensation is separated out."
On the right track
Daryl Innis, vice president and practice leader, communications components, at market research and analysis firm Ovum-RHK (search for Ovum-RHK), thinks the companies are wise to team up. "I think it's a good thing for each of the two. JDSU didn't seem to really have a transmission module that was 40-gig ready," he explains. "I think Mintera needs financial backing, and partnering with JDSU also gives them some credibility."
He also likes ADPSK. "What they've developed can be considered a very attractive approach," Innis says. "A lot of the carriers and the network equipment manufacturers want 50 GHz -- or at least they want it in their back pocket. They have the reach [with ADPSK] and you don't have to pay an awful lot more; you're paying about the same as what you pay for DPSK. So you don't have to go to DQPSK and other modulation formats."
Innis says that he sees "a significant opportunity" in 40 Gbits/sec, which he described as being in an early commercialization phase. For this reason, the partners should be well timed if they get their module into the market on schedule next year. Innis says that early competition should come from StrataLight, Opnext (which demonstrated a 40-Gbit/sec module at ECOC), and Optium (in particular via its acquisition of Kailight). Other vendors who should not be discounted in the space include Finisar and Yokogawa (which has focused on DQPSK).
Innis says that DQPSK is the modulation format of choice in the Japanese market. He suggests that module vendors looking to enter the 40-Gbit/sec market face a critical choice when it comes to choosing which of the available modulation formats to support. Iwata says that while the immediate co-development work with Mintera will focus of ADPSK, that doesn't mean the company wouldn't offer products that support other modulation formats.