Sunday |
|
|
| Interview |
Interview with Stephen King, Chief Executive Officer.A founder of Phyworks, with 20 years high-tech marketing and sales management experience, Stephen King spent the previous nine years in the fibre optics industry at Maxtek Components as Business Development Manager for fibre optic transceiver products, VP of Business Development at Microcosm Communications and VP of Marketing and Applications at Conexant Systems in the then ICON (ICs for Optical Networks) product group. Company overview Phyworks, a privately held company founded in 2001 and based in Bristol, UK, is a developer of high-speed communications chips designed to significantly cut the cost of 10 Gbit/s and below copper and optical interconnect. Phyworks uses proprietary algorithms and high-speed mixed-signal design to create chips that compensate for the optical characteristics of fibre, such as dispersion and noise, as well as copper cables and circuit board traces and connectors in standard mixed-signal CMOS devices. The company sells its products into telecom - specifically FTTH - and datacom markets worldwide. Phyworks has a substantial presence in the PON market, with customers in the U.S. and Asia, including China, Korea and Japan. Introduction Phyworks started out developing ICs to enable customers to increase the capacity of existing fibre infrastructure, with a focus on upgrades of 2.5 Gbit/s metro networks to 10 Gbit/s. Explained Stephen King, "Phyworks' initial product addressed Forward Error Correction (FEC) to deal with attenuation over optical links, however it became clear that customers also needed to manage dispersion, so the company then developed an Electronic Dispersion Compensation (EDC) chip". Subsequently Phyworks developed products for the datacom market based on its existing EDC solution and a Physical Media Device (PMD) product line with applications in both the telecom and datacom markets. The company currently addresses the Passive Optical Network (PON) market, offering a range of transimpedance amplifiers (TIAs) and laser drivers / post-amplifiers, and backplane and active copper cable markets with its equaliser and clock and data recovery (CDR) products. Products and technology Mr. King noted that following the telecom market crash of 2001, the industry did not recover to the level expected, and by the time Phyworks' products were ready commercially in 2003 it was clear there was no equipment into which its FEC and EDC chips could go. At this point the FEC product was shelved and the company turned its attention to the datacom market. Specifically, Phyworks redesigned its EDC product for the long-reach multimode (LRM) market, which was then being pushed by Cisco in particular. Stephen King said that it soon became clear the problem Phyworks was seeking to address was highly complex, with numerous companies - both start-ups and established players - trying to solve it. Again, the market for the product did not materialise to the extent anticipated. However, Phyworks had also commenced work on a second technology with its PMD line. These chips are targeted at applications in optical modules for both the datacom and telecom markets. Mr. King said that Phyworks key differentiator in this market was the level of feature integration it could offer, adding, "For the first PMD chip customer Phyworks integrated what had previously been three chips into a single-chip solution. The device was supplied at the same cost as a three-chip product but offered advantages to the customer". It was noted that Phyworks' first PMD chip customer used the device in a Fibre Channel (FC) system that addressed the shift from 2 to 4 Gbit/s FC technology. The FC market is highly competitive, and when Phyworks introduced its first product many optical module suppliers were selling their products into the market at a loss. Mr. King noted that Phyworks has always focused on delivering products that reduce power and space requirements and overall cost for the customer. Phyworks is also continuing development of its equaliser products based on an analogue/mixed-signal implementation and integrated CDR, as well as new continuous time equaliser technology. The company offers its equalisers with different interfaces for IC solutions with applications in a range of markets. According to Stephen King, the company intends to offer its equaliser and CDR products via a mix-and-match approach to address different applications. The chips are currently supplied to the backplane and active copper cable markets, but there are numerous opportunities in the active copper cable sector, ranging from HDMI links to data centre infrastructure. Concerning Phyworks ability to produce multiple products for a range of markets, Mr. King said that developing a number of different products and attaining qualification for new products presents a challenge for an operation with limited financial resources and only 40 staff. Phyworks currently holds 11 patents and is in the process of filing for a number of additional patents relating to its latest generation of PMD and equaliser technologies. Patents need to be held in the U.S. - an expensive procedure - to be of real value, as a UK patent will not protect IP if a competitor is based, for example, in the U.S. or Asia and sells its products in markets other than the UK. Transition to CMOS Phyworks is in the process of transitioning its PMD product line to CMOS process technology and has so far developed two CMOS-based products with a third preparing for fabrication. Mr. King said the technology migration is proceeding smoothly, noting that the first product is a 2.5 Gbit/s TIA and the second a short range transceiver. Regarding the TIA device, Stephen King said the product is now moving into production after sampling and has demonstrated the best performance in the industry, including against Phyworks own SiGe-based device, which was previously claimed to be the best performing product available. It was noted that the 10 Gbit/s SR transceiver product has now been delivered from fab prior to going into commercial production. Mr. King stated that Phyworks has considerable experience in CMOS as all of its equaliser products are already based on this technology, adding that the issue with CMOS until recently has been cost. Specifically, for a 10 Gbit/s device, the minimum requirement is a 130-nm CMOS process and until recently it was cheaper to use, for example, a 0.18-µm SiGe process. However, the cost of using 130-nm CMOS has now dropped to the point where it is competitive with SiGe process technology. Market focus Now focused on the PON market, Phyworks has been successfully supplying products for the ONUs and MDUs installed at customer premises, with many network deployments in the U.S. and almost all in Japan utilising Phyworks technology, according to Mr. King. In addition, Phyworks has a strong presence in Korea in this sector and has also made inroads in China during 2009. "Phyworks gained an estimated 50% share of the PON market in 2009. Aside from straightforward survival, the company aimed to increase its position in the PON space last year and was successful, marked by its progress in the Chinese market". Worldwide, Phyworks estimates around 7 million ONUs were installed in 2009, while the market is forecast to grow to between 13 and 14 million units by 2012. With respect to Phyworks ability to maintain its share of the PON market, Mr. King noted that there are several factors that will affect the market over the next few years, a major one being the move from B/E/GPON to 10 Gbit/s PON/EPON that is expected to start from 2011. Phyworks is currently developing new chips for 10 Gbit systems, although Mr. King expects the shift will give competitors an opportunity to enter the market, adding, "To date the company has held its position with respect to the PON market; Phyworks has not lost any accounts over the past three years. This achievement is based on offering the highest level of integration and overall lowest cost per deployment". On a geographic basis, it was noted that Phyworks does around 50% of its business in the U.S. - with Asia the primary market for PON solutions globally. On the equaliser side of the business, systems design work for the telecom market is still carried out in Europe, notably Italy and Germany, and there is a market for the equaliser products there. Stephen King pointed out that Phyworks has recently signed up a customer in Germany. Market share Phyworks claims to already have around a 50% share of the PON market and Stephen King acknowledges that expanding through increased share in that sector will be difficult but he does expect to achieve some growth by that means. However, most sales growth in the PON sector will be due to overall growth in market size, forecast at approximately 20% per year over the next few years, as well as the growth of 10G PON where Phyworks will introduce CMOS transceiver product during 2010. The company expects to leapfrog the competition through higher levels of integration in cost-effective CMOS process technologies. However, Mr. King stated that Phyworks is directing significant investment into other areas, primarily its equaliser technology. He added that Phyworks presently ships in excess of 50,000 units per quarter in this sector and recently introduced a new high performance, low power equaliser product for a new market segment. Pricing and margins Regarding profit margins, Stephen King said that while on the PMD side profitability is tight and the company is constantly seeking to reduce its costs, on the equaliser side Phyworks' products are high-end with more value and so margins are better, and reflective of the higher development costs. In the PON market, Mr. King noted that customers are constantly seeking to lower costs as carriers tend to almost give away ONUs to encourage subscriber take-up. This pricing pressure is passed on to the carriers' system vendors, and in turn the chip suppliers. He added that in the Asian market, module suppliers typically expect prices for their products to fall by 20-30% per year. Stephen King believes that ultimately it is down to the suppliers to choose whether to stay in a particular market. Suppliers can usually make viable profits on a product early on and if it moves into 'cash cow' mode, before the product finally becomes uneconomic to produce. At this time the supplier must decide whether to stop offering the product as it focuses on the next generation. However Stephen King remarked that the PON market has yet to reach the latter stage. Given the relatively small size of the PON market, it might seem surprising that there is such pricing pressure but Mr. King explained that this is principally due to the number of Asian suppliers in this specific sector, with almost all manufacturing taking place in Asia and most of the new entrants to the market being Asia-based and seeking to win market share by offering the lowest price. Mr. King said a core issue in the optical industry is that many suppliers now have only manufacturing capabilities, whereas in the past companies tended to be module suppliers, adding value to their components. Today the bulk of manufactures in the sector do not have the optical capability and aim to compete on price alone. As a private company Phyworks is not required to publish financial data, although Mr. King remarked that the company does not seek to hide such information and any potential customers will see its financials as part of the due diligence process. He noted that the company's revenues will grow by 45% year over year for the current fiscal year ending June 30, 2010 and 60% overall for the full calendar year. Competition Regarding competition in the PON sector, Stephen King considers that a major factor behind customers using products from other suppliers is historical, particularly in China. Another factor cited was the lengthy approval cycles for new products. Despite a general reluctance of customers to change supplier, they also usually want a second source for products so new suppliers do have opportunities. Mr. King added that while it can be very tough winning business with a new vendor, once won, the supplier is likely to keep the business unless they doing something wrong or prove to be an unreliable supplier. Certain of Phyworks' competitors have an advantage in terms of size and based on a purely financial comparison a customer might choose a competitor over Phyworks. But Phyworks seeks to overcome this by providing more highly integrated solutions at cost competitive prices and lower power, backing this up with what it considers the best applications, product delivery and sales support in the industry. This has thus far proven to be a successful strategy for the company. On the subject of due diligence, it was noted that as yet the process has not extended as far as the end (carrier) customer for a product utilising Phyworks chips, which could encompass companies such as Verizon. However, Mr. King stated that such a procedure could become a requirement and is always a possibility. Phyworks has been audited by a number of Japanese and U.S. customers and has not lost a sale because it could not pass due diligence - it has always been able to prove the viability of its business - and being profitable helps a small company significantly. Mergers and acquisitions Considering Phyworks relative lack of financial strength versus larger players, Mr. King acknowledged that by this measure it could be beneficial if Phyworks were acquired by a larger company, noting that it has been approached, "Phyworks is successful and currently growing so it is clearly a potential acquisition target. Particularly over the past few months, semiconductor companies have been seeking to implement acquisitions in the communications industry. As yet Phyworks has not accepted any offers and is keen to expand organically". Commenting further, Mr. King said that among other advantages, being part of a larger operation would provide access to new markets and customers and the fact that the due diligence that must be completed with customers, particularly systems vendors, is a significant commitment for a small business such as Phyworks. However, he emphasised that being acquired is not a goal at present, although he would not rule out the future possibility in the right circumstances, or even perhaps a strategic investment. In terms of itself acquiring or merging with another company, Mr. King said that one reason for not doing so as yet is the risk involved, although the company has been looking to engage in such a transaction over the last year. He noted that any acquisition has to be based on very careful consideration of the drivers behind it and a close assessment of how it will affect the employees involved, "A key element to any merger or acquisition is the staff, as they will certainly be a very important part of the acquired business - typically more so than the products". Mr. King did reveal that Phyworks is looking at potential M&A targets in the Bay area, and to support that activity set-up an office in Santa Clara in October 2009. Scale is important Mr. King believes and feels there is no future in the long-term for one- or two-product businesses in the sectors where Phyworks operates. He further noted that actually implementing a merger or acquisition has proved difficult, particularly where two private companies are involved, "The main issue relates to valuing a business. Phyworks is profitable but some companies that it might consider combining with are not. Nevertheless these businesses, often backed by VC money, believe they hold significant value. Such companies would prefer to be bought by a public company as this will bring liquidity, which would typically not be the case with a private buyer". As to the type of business that Phyworks would aim to combine with, Stephen King said that as an analogue/mixed-signal chip supplier it would be primarily interested in integrating its technology with digital IP, perhaps a supplier of media access controllers addressing a different target market. Another attractive option would be to combine its technology with high-speed analogue IP for adjacent markets to those Phyworks currently serves. Mr. King emphasised that Phyworks core expertise is in ICs and the company would not be interested in moving up the chain into, for example, the optical module business - a sector generating lower margins but higher revenue than the chip business. Equally, he said, Phyworks has not been approached by an optical module supplier regarding a possible combination, likely due to that fact that there are numerous chip suppliers and therefore there is limited value in such a move. Considering the alternatives for Phyworks, Stephen King said that failing the completion of an acquisition or merger, the company would aim to grow organically with a view to launching an IPO once it achieves sufficient size. The only other available option would be to accept a buyout offer from another company. The road ahead As to the future for Phyworks, Mr. King sees the biggest threats as the economy, competition - for example if a larger rival decided to go all-out to try to take Phyworks business - and financial factors, particularly relating to the company's ability to fund multiple product lines from cash flow. A further financial concern would be the failure to pass a potential customer's due diligence requirements. While customers are impressed on technical side, Phyworks is a relatively small operation with limited financial resources. A customer reviewing Phyworks' financial position would find an efficiently run company that has reached its current position with only $32 million of funding, "I would be surprised if there is another company in the communications sector that has done as much as Phyworks with so little funding. We've passed financial due diligence with a number of large Japanese and U.S. companies. I believe the fact that we are efficiently run and profitable helps immensely". Phyworks' most recent funding round was completed in 2007 and the company is now financing its operations from cash flow. However, Stephen King did say that the company might seek additional funding to finance a specific objective, such as accelerating the development of a new product line. The company does have a number of strategic investors interested in looking at possible funding but valuations are always a struggle in the current market, "Still, if a strategic investor could bring more to the table, access to different markets, customers or new channels to sell products, we would be interested in a discussion". Mr. King commented that its investors have seen the performance of the company through some very difficult economic times since 2001 but especially recently Phyworks has produced and achieved significant growth, "We were named as Great Britain's 5th fastest growing private technology company, and the #1 fastest growing semiconductor company in the 2009 TechTrack 100 sponsored by the Sunday Times and Microsoft. That buys a lot of credibility". Concluding, Stephen King said in 2010 the company is aiming to expand through the addition of sales, applications and marketing staff in the U.S., to build its presence in that market, while maintaining development staff, finance and operations in the UK.  |
|||||||||
