BLOG

Absurdity of Stressing Metro 100G Now

January, 2015

A couple of financial analysts have been touting Ciena as the expected big winner for a future contract for a large 100G metro project at Verizon. While there can be little doubt that the vendor would be part of such a deal, it needs to always be remembered that these announced agreements never compel the service provider to buy one dime from any supplier, and so the actual amount of equipment purchased could be far less than initially indicated. Moreover, the conspicuous absence in the current discussion of Fujitsu, a vitally important metro incumbent vendor to Verizon, strongly points to the service provider buying less than a stellar amount of this gear anytime soon. Above all, current expectations for growth in the total 100G market are nothing to write home about – one market research firm, usually among the bullish crowd in general, is only projecting about a 25% increase in terms ...

Read More

Google and Cars: Same Revenue Stream

December, 2014

There tends to be a lack of a full understanding of the intention behind the bulk, if not all of Google’s investments in new solutions. At the end of the day, they are each about the enablement of, or at the very least, showing the way of increasing the amount of Internet traffic in order for the company to enlarge the size of its principal business – advertising on the web. Google’s work on the self-driving automobile is no different in that to whatever extent a driver can legitimately divert attention while on the road, that person can potentially be spending time on the Net. In comprehending the mentality behind Google’s investments, one can see why it is pushing for the ultimate vision in autonomous automobiles. It is inconceivable that the firm would want to actually be a player in such a heavily regulated and extremely competitive market as well as open itself up to an incredible a ...

Read More

Cisco: Buyer of Potential Finisar-JDSU Combo?

December, 2014

If Finisar purchases JDSU’s CCOP division, would a systems integrator be interested in buying the combination? Unquestionably, there have been rumors in the past that Cisco Systems would buy Finisar, which with CCOP, would make Cisco’s previous component acquisitions pale in comparison in terms of cornering the market. Moreover, the full system supplier has been Finisar’s biggest customer for a long time. There is a school of thought that when Cisco decided to cease having a distinct transport business unit, it was making a statement that optics stopped being an end in itself – it was really a means to an end. While selling optical network gear remains part of its general strategy, its principal purpose is supposedly to tie routers and the server farms as well as offer the backbone for its software. According to this theory, Cisco’s purchases of CoreOptics and Lightwire are more evidence ...

Read More

Finisar Buyout of JDSU’s CCOP Inevitable?

December, 2014

Simply by a process of elimination, it is difficult to foresee any other player besides Finisar being interested in purchasing JDSU’s planned spinoff of its optical components portfolio, Communications and Commercial Optical Products (CCOP). Such a deal would provide Finisar with further diversification in the telecommunications transport space as well as potentially in new business sectors. Also, to whatever extent they presently go head to head with each other, it would eliminate a competitor from engaging in the cutthroat pricing behavior that is so prevalent in the market today. With its heavy focus on data communications, some industry observers have mentioned Avago Technologies as a possible acquirer of those JDSU assets in order to broaden its product line. However, there is a high likelihood that Avago would have a hard time justifying such a takeover after its relatively recent ...

Read More

Data Centers Moving to “Sweat” Shops

December, 2014

What is really happening in the greening of the data center? We are not seeing orders of magnitude reduction of power consumption. We are not seeing the use of cleaner power plants. Evidently, engineers and other individuals working at Data Centers (DCs) will be increasingly making wardrobe changes from light jackets and slacks to shorts and sleeveless shirts. The expectation is that the standard operating mode for many DCs will be as high as 90 degrees Fahrenheit. At least one major components vendor is working on chip sets, which will work more effectively under such conditions and actually, the permitted temperature for a server can normally reach 90°F (although it is usually advocated not to go beyond 77°F). Obviously, in moving in this direction, the cost of cooling a DC can come down substantially. However, what is not as recognizable is the extremely high cost of having all of tha ...

Read More

New M&A Types of Players In Optics

November, 2014

In recent weeks, there has been a flurry of acquisition activity with three optical component companies being purchased by non-traditional kinds of acquirers, particularly players in the short distance, communications market including data centers. These buyers, such as Huber+Suhner and its takeover of Cube Optics, are definitely concerned about an adequate supply of componentry being available. Evidently, even big enterprises with large data centers have the same types of worries as they are also investing in optical device development, as pointed out in a previous blog article. M/A-COM bought BinOptics at the premium price of $230 million (probably four to five times revenue) to ensure adequate access to laser semiconductors. The bottleneck created by Avago Technologies' pickup of CyOptics last year was a significant catalyst. In the past, JDSU and Finisar had a major role in such M&A ...

Read More

AT&T’s Neutrality Reply a Red Herring

November, 2014

While on the surface, it is commendable that a large player such as AT&T is taking on the US government on net neutrality, its response is totally self-serving. The carrier knows that it will make relatively little money in providing fiber to residential customers. Plus, the biggest cause of network congestion does not come from uploads and downloads to and from the home; it is about the interconnection of data centers for high-volume content. Verizon’s concentration with fiber to the home was in areas that had large enterprises nearby. It divested a lot of its other lines that did not fall into this category. Verizon saw an opportunity to disguise aggressively going after the very lucrative business customers from the regulators – and FTTB subsidized FTTH. To this day, lots of people in the industry would prefer to believe Verizon was just crazy to do so much FTTH, especially several ye ...

Read More

Telecom Analytics With Bad Data?

November, 2014

On the landline side, the incumbent service providers have been notoriously lousy at record keeping. There are still in place very antiquated inventory systems along with a high level of ignorance about whether a large number of circuits are either working or connected in networks. At least with wireless technology, it has been around for a much shorter period of time, and would not be nearly as burdened with ineffective solutions in the back offices. What is the biggest reason for bandwidth on demand not taking off in a significant way after about three decades of hype? Again, it is about the lack of adequate knowledge by these carriers concerning their networks. How can executives be sure that rapid change will not adversely impact services with them all toppling on to each other? While there have been workarounds, modifications, as well as reductions in the use of Ma Bell’s Trunk Inte ...

Read More

Net Neutrality: Hurtful Government Whims

November, 2014

The notion of neutrality in general is a myth. Nations that claim to be neutral in inter-state conflicts are hardly impartial. If country A has a major advantage in its conflict with country B, the so-called non-aligned party is in effect supporting the former, and potentially leading to its winning the fight. When the US government is pushing the notion of net neutrality, it is also a canard because corporations by definition are not neutral because they want to defeat their competitors and make money for their shareholders. Arbitrarily forcing them to provide bandwidth in ways that are not economical will only suppress new investments in technology – leading to even more constraints in networks. The capricious nature of the US government to decide on policies can have beneficial results. The original breakup of AT&T in 1984 led to a massive fiber optic build-outs because MCI and Sprint ...

Read More

Integrated Photonics Manufacturing Institute: “Panoptic” Danger

October, 2014

One should hardly rule out the risk of the Integrated Photonics Manufacturing Institute (IPMI) morphing into an all-encompassing and permanent industrial policy for the US. The term “integrated photonics” is so ambiguous and nebulous as to potentially include every facet of new optical technology. Also, it is a sure bet that manufacturing itself will not turn out to be the essential role of IPMI in the long term, rather actual R&D work on newfangled solutions will be its focus – with the potential of hampering future technical development far into the future. There are ample precedents for private-public partnerships straying away from their original missions. SEMATECH’s essential mandate at the beginning was really just to increase knowledge of advanced production processes for semiconductors. Later, the consortium’s goals changed to include branching out into the build-up of infrastruc ...

Read More

Uncle Sam: Major VC for Optical Gear Development Indefinitely?

October, 2014

Where will the money come from to fund R&D for next generations of fiber optic equipment in the telecommunications and data communications markets? Clearly, the vendors will have enough trouble just keeping up with the changes on existing equipment in what are many cases relatively low-margin businesses – and will be quite hesitant to move ahead significantly on next-gen devices before they get an adequate return on their original investments, including 100G. Despite the heavy amount of cash going into silicon photonics, the increasingly pessimistic outlook for its future could easily result in the last straw for venture capital firms for an extremely long time, which almost assuredly to this day have not fully received their money back on devices at much lower speeds, including those systems operating at 10 megabits and at 10 gigabits. Certainly, the situation of getting an adequate amo ...

Read More

OpSIS’ Demise: Politically Incorrect Si Photonic Wafers

October, 2014

This past July, OpSIS announced that the program funding its core operations for silicon photonic Multi-Project Wafers (MPWs) had ended and that it would cease operations. Evidently, the US government around that time was uninterested in supporting the continued existence of the foundry, which had been receiving all of its capital from the private sector. Only about a few months later, the Administration along with the Department of Defense (DOD) apparently changed its mind on the importance of such MPWs, as the White House came out with its $200 million initiative – in effect, making OpSIS a loser before the Integrated Photonics Manufacturing Institute (IPMI) was even established. A fundamental problem with public-private partnerships is that it not only puts the government in the position of picking the winners of technological solutions, but the actual players involved in the game. Wh ...

Read More

Intel Behind $200M Photonics Competition?

October, 2014

The US government did not call it a “silicon photonics contest.” Perhaps the federal funding for this new Institute for Manufacturing Innovation (IMI) would permit Intel and others to gradually shift to more of a generic term for its future development of chips that includes optics, but would not necessarily involve silicon. Actually, in a recent announcement, IBM implied that its new $3 billion, five-year plan for chip development, might not comprise Si photonics. IMI is quite reminiscent of SEMATECH (SEmiconductor MAnufacturing TECHnology), incorporated in 1987, which became a not-for-profit consortium that included the US Department of Defense (DOD) as well as semiconductor suppliers and educational institutions. Its main purpose was to grab back leadership on market share with chips, which was taken away by Japanese companies. There seemed little doubt that early on, Intel was runnin ...

Read More

A Kaiam Strategic Partner: Google?

September, 2014

Concerning the recent $35 million announcement by Kaiam of equity funding, our intelligence indicates that one of the two strategic partners is Google. We also think the other partner is Microsoft. In addition, we believe that the total valuation is two to three times revenue -- so over $70 million. While Kaiam has clever packaging, we are not convinced that it has the lowest cost in the market. Nevertheless, the supplier has benefited from getting out there early in the data center/datacom space. It is likely that it owns the 10x10 business in that sector. Unquestionably, signs of two giant enterprise firm funding Kaiam may be about ensuring direct access to unique 100G technology. Perhaps it is the vendor’s Hybrid Mixer PLC – and/or maybe something else that could possibly be on the drawing board that involves a coherent, serial 100G device. More practically speaking, there is a higher ...

Read More

Videoconferencing: The Elusive Business Application

September, 2014

Although it was 50 years ago that AT&T introduced the Picturephone at the World’s Fair, the company is still striving to “encourage the adoption of videoconferencing” at major corporations today. There has just not been a significant cultural shift in the business community to making deals without shaking hands in person. Yet, there are also very practical considerations in videoconferencing providing an inadequate sense of being in the room, and not allowing for a close consideration of body language to assess the impact of a presentation. While webinars have undoubtedly reduced the amount of traveling substantially in prequalifying sales prospects in offering graphical video displays in communications, there is no real need to have a live picture of the people in the conversation. So, videoconferencing is partly about using a solution in which there is not a problem. Also, somewhat iro ...

Read More

Finisar Getting a Raw Deal with Silicon Photonics

September, 2014

When it comes to technology development, some of the leading analysts in the financial community have a tendency to blindly follow a template of conventional wisdom. One of the most egregious examples was during the bubble when there was the widespread notion that the need for bandwidth could not be satiated, and service providers were literally being punished for installing insufficient dark fiber. A similar injustice has occurred with Finisar with the imagined threat to the company by the introduction of silicon photonics. These investment firms talk about the “checks” they conduct in drawing their conclusions. Well, how about spending approximately five minutes checking with an objective optical scientist? As we suggested in a recent blog article, it is clear that in general, development of silicon photonics for active components has for several decades been akin to placing a square p ...

Read More

Silicon Photonics: Widespread Usage on Active Components Doubtful

September, 2014

While the amount of industry chatter about silicon photonics rose substantially in 2014, partially because of certain market research firms looking to sell more reports, the technological hurdles, especially with insertion loss, make extensive use of active, combined components, such as for modulators and photodetectors, unlikely in the foreseeable future. Although there have definitely been numerous demonstrations of integrating with these CMOS fabrication processes, like with receivers, both the performance and cost are superior with standard methods. Certainly for passive devices, silicon lends itself quite well to integration, including for AWGs. Instead of using micro optics, combining waveguides with attenuators is being accomplished with Si. With 100G receivers, the delay lines are being integrated with the phase combiners with either silicon or Indium Phosphide. Given the increas ...

Read More

The Guts of the CFP2

August, 2014

For the next generation of 100G coherent line interface form factor CFP2, will it be a split implementation -- analog CFP2 and separate digital DSP ASIC for the card? Can both parts be integrated into the single package? As far as we know, it will be split. While today, there is one 100G on a card, etc., with CFP2, there will be at least two on a board. There is no question that it would be desirable to use the same DSP for both. However, the industry has gone back to the 1990s in which 100G customers, such as Alcatel-Lucent, Cisco Systems, etc., each have their own DSP, and they believe that to be their secret sauce -- and that is why the DSP is coming out of separate packages. (written by: Alka Swanson)

Read More

25G Laser Shortage

August, 2014

We have been told that there is going to be a worldwide shortage of 25G lasers in the coming two years, supposedly due to not enough capacity having been planned by the industry as a whole.  We hadn’t heard anything like this before.  Do you guys know if this is true, and if so, why it happened? A couple of our contacts indicate that supply is down because of unexpected demand by a few data center operators as well as a greater need for customized optics, and that 25G lasers have inherently lower manufacturing yields than is the case, say with 10G devices. Specifically, the 25G DMLs are in short supply, not the EML devices. However, the EMLs are very expensive, especially for short-distance applications. There is also apprehension that expanding production will create a long-term overcapacity situation, once the market gets back to normal levels. In the meantime, it has become a waitin ...

Read More

Big Need for MPLS-TP?

August, 2014

At the service providers across the globe, the most widespread and native selection is MPLS and not MPLS-TP. The former has broader acceptance, superior interoperability, and operational intimacy for the engineering folks. In Asia, particularly China, MPLS-TP has taken off in a big way. While some industry observers view that part of the world as exceptionally unpredictable and may suggest that “TP” could even be replaced, it seems that at least one of the large Chinese operators has such a critical mass of deployment that such a move is not imaginable. On its face, one would think customers should want to go with MPLS-TP over MPLS because the former is a newer version. Ordinarily, any supplier that would happen to start product development from scratch would definitely be inclined to begin in the “TP” direction. Theoretically, MPLS-TP should be a slam-dunk in a greenfield situation at ...

Read More

Impact of Cisco Systems’ Purchase of Tail-f?

August, 2014

What happens to Tail-f and to all of its customers after the acquisition by Cisco? Do those customers have to find a new partner? Will Cisco keep it open to all vendors and competitors? Does Cisco imbed it and no longer allow access to third parties? Each customer probably had an agreement with Tail-f in case of an acquisition. It may be on a case-by-case basis, but in terms a macro answer, these days, there are not purchases without putting in a clause for M&A. For small suppliers, it can be quite detrimental. For example, regarding a component company that was purchased in 2013, large module and system suppliers, such as Finisar, Juniper, etc., put in a stipulation that if the vendor was acquired, and did not supply the devices at ridiculously low prices, then they can take hold of the IP, and produce them somewhere else. It is a Catch 22 situation. If one refuses to sign such a contra ...

Read More