Movers & Shakers Interview with Eran Eshed - Frost & Sullivan

Movers & Shakers Interview with Eran Eshed
Frost & Sullivan
Luke Thomas

March 4, 2009
   
In initiating an exhaustive review of the key drivers and challenges for next generation wireless technologies such as Mobile WiMAX, 3G LTE and XGP, Frost & Sullivan interviewed the Co-founder and VP of Marketing and Business Development, Eran Eshed of Altair Semiconductor, who is making significant traction in the mobile broadband wireless access market. In this interview, Luke Thomas, ICT-Europe Programme Manager for Mobile Broadband Wireless Access (MBWA), discusses some of the keys issues pertaining to mobile broadband technologies, Altair Semiconductor's sustainable competitive advantage in their service offerings and the impact of the economic downturn in the mobile and wireless industry.

Luke Thomas: Altair Semiconductor has been receiving a lot of attention in the press lately. For those not familiar with Altair Semiconductor, can you provide a snapshot of the company's activities?

Eran Eshed: Altair Semiconductor is considered to be the world's leading developer of ultra-low power, small footprint and high performance 4G semiconductors. Our goal is to develop the highest performance, lowest power consuming and cost effective 4G chipsets that will enable connecting any device - be it a notebook computer, a mobile phone or a music player - to the Internet. Our mobile WiMAX and XGP chipsets today are the industry's undisputable benchmark for ultra-low power consumption and very small footprint, a merit which our LTE chipset will stand out for as well. Our 4G product portfolio includes not only baseband processors but also RF transceivers for TDD and FDD applications, and is arguably the most comprehensive 4G semiconductor portfolios available from one company.

Luke Thomas: Please share with us what your current role at Altair Semiconductor entails and what you did prior to this job?

Eran Eshed: I am responsible for marketing and business development at Altair, and am one of the three founders of the company. My founding colleagues and I were part of a startup company called Libit Signal Processing, which was acquired by Texas Instruments in 1999 for $365 million. Our team became the core of TI's Broadband Communications Group (BCG), which developed products in various segments such as DOCSIS cable modems, DSL, WiFi, GBit-Ethernet, DTV and others. As part of TI's BCG, I was responsible for marketing of the cable modem product line. My professional background originates in electrical engineering, hardware and chip design where I spent the good part of 10 years of my career.

Luke Thomas: Has the global recession created any dramatic change in Altair's product roadmap of a multimode WiMAX / XGP/ 3GPP LTE silicon solution?

Eran Eshed: No, the recession has had no impact on our product roadmap. From day one, our product strategy has been to develop multi-mode 4G semiconductors and we are executing on it well. Our commitment to WiMAX remains while we are getting very close to rolling out our multi-mode WiMAX and LTE chipset. I see this as a major differentiator for Altair. As the industry shifts and changes, other companies in the space are in the process of shifting their roadmaps to align with it - mainly to add LTE. For us, this is not a shift in the strategy. Rather it IS the strategy and has been so from the company's inception. In addition to that, we are well funded and closed a large $22M round in September of 2008. This, in the addition to strong revenue growth we experienced in 2008 and which we expect to continue this year, positions us well to continue investing in what we believe is strategic to the company's future growth, as opposed to having to cut back on whatever does not generate revenue today.

Luke Thomas: How does Altair differentiate itself from competitors such as Wavesat?

Eran Eshed: On a general note, Altair is really the only semiconductor company that has been investing in both WiMAX and LTE in complete parallel for the past 2.5-3 years. Our first chip specification dated August 2005 defined full compliance with mobile WiMAX, 3GPP LTE and anecdotally 802.20 (later to become the basis of Qualcomm's UMB which failed to gain momentum). Obviously this required a great amount of programmability and flexibility, however in comparison to 2G or 3G technologies, the processing power required to process 4G signals is dramatically higher, so the usage of traditional DSP based architectures yields a product that consumes large amounts of current and drains battery very quickly. The way we approached this flexibility vs. power-efficiency paradigm, is by developing a software defined 4G/OFDM processor which we codenamed O2P (Optimized OFDM Processor), and which outperforms any conventional DSP being used today in WiMAX or LTE designs by more than an order of magnitude. So we have exactly the right amount of flexibility required to support practically any 4G standard - yet without compromising power efficiency at all – the O2P stacks against any optimized ASIC gate logic on processing performance and power efficiency. In October of last year Altair was announced winner in the “Mobile WiMAX Low Power Shootout” which took place in Chicago, in which we beat our closest competitor with factors ranging between 4 and 10. We see ultra-low power as our most significant differentiator, especially when combined in a multi-mode 4G architecture. In addition, our chipsets are the smallest in the industry in both WiMAX and XGP, and will continue to maintain this advantage moving forward into LTE.

Specifically in reference to the competitor mentioned in the question, on the surface it may appear that we have similar product and market strategies – the big difference however is in the underlying technology and in the depth of the portfolio. From the technology aspect, the architecture from the company you had mentioned, as are most implementations in the market today, is based on a traditional multi-DSP core. In this case, DSPs that were originally architected to process low-bandwidth audio signals, and were enhanced and combined together to increase the processing capability. Altair's O2P is far superior to any DSP implementation in the market, as I believe we had proven beyond doubt in the WiMAX space. Aside from power consumption, cost is a very significant parameter in high volume markets such as we expect 4G - especially LTE - to become, and it is a fact today that our die size is by far the smallest for both WiMAX and XGP and our CMOS process is more streamlined/common and hence the wafer costs are lower. So there was a great amount of thought and effort put into developing for low cost, and we think we have a significant advantage there as well.

In terms of the depth of the portfolio, I strongly believe that companies without RF transceiver products will find it extremely difficult to compete in LTE. Companies, such as the one mentioned, could make it in WiMAX as they used RFIC's from pure-play RFIC companies such as Maxim or PMC Sierra. The situation in LTE is completely different, as such RFIC companies understand that this is a game of the large cellular chip companies and will not offer such products to the market. Altair's RF product line covers transceivers for both TDD and FDD in most LTE bands and markets – so this is a dramatic competitive advantage for us.

Luke Thomas: Is 3G LTE comparable to Mobile WiMAX with regard to performance characteristics? Kindly explain.

Eran Eshed: The short answer is yes, because they are both based on similar fundamental underlying communications concepts – the modulation which is OFDMA in the downlink, the signals bandwidth which can go up to 20MHz, the usage of advanced smart antenna techniques, to name a few. LTE as defined today is a little more advanced and is capable of supporting higher theoretical throughputs. However, the next generation of WiMAX will close this gap and may even leap ahead. In terms of capacity, cost per bit, spectral efficiency, noise immunity etc, no major differences exist.

Luke Thomas: When will certified devices complying with Wave 2 Phase 2 be out in the market? Kindly specify what the client devices will be –USB Dongles, UMPCs, smartphones, laptops or mobile devices.

Eran Eshed: I expect to see certified products towards the end of this year. Unfortunately, the only devices that are capable of being certified today are PC and notebook peripherals, or CPEs. The certification program theoretically supports certifying embedded handheld devices, however none have yet gone through it and I expect it to take a little more time than the dongles.

One of the challenges faced by equipment vendors today is that even after you certify your device, there is a good amount of additional testing and interoperability that has to be performed in the specific carrier network in which you target deployment. This causes the time to market and cost of bringing devices to market to increase, and I hope that we as an industry will find the ways to improve the process and lower the barrier to certification and revenue generation.

Luke Thomas: Do you see 3G LTE being ratified as a standard by March 2009?

Eran Eshed: I hope it will, however this is more of a formality since the standard is very stable and change requests here and there do not change anything fundamental. From a semiconductor perspective, there is absolutely no issue freezing specifications and taping out chipsets. The software requires tracking and updating, but this is a very natural process in the formation of a new standard.

Luke Thomas: Due to the economic downturn, do you see mobile operators fully maximizing their existing HSPA/EVDOx networks before deploying 3G LTE?

Eran Eshed: Generally speaking, yes, it is clear that operators who have made significant investments in existing infrastructure will try to maximize the return they get on it. HSPA seems to have a stronger upgrade roadmap than EVDO, but both still offer upgrades which despite not being as significant as LTE in terms of performance improvements, cost less to deploy. The three leading carriers today in terms of LTE deployment – Verizon Wireless, NTT DoCoMo and China Mobile - each have their own reasons for accelerating LTE deployment, and at least two of them have very significant amounts of cash to execute on these plans. Overall, I expect to see 2010 as a major trial year for LTE and 2011 for real commercial service launches.

Luke Thomas: Will 3G LTE have an advantage over Mobile WiMAX with regard to better battery life of client devices as 3G LTE uses SC-FDMA in the uplink?

Eran Eshed: SC-FDMA does indeed allow usage of higher efficiency power amplifiers, and hence provides LTE with an advantage, at least on paper. However, the whole power consumption aspect should be looked at from a system level perspective and is greatly dependent on parameters such as cell density and size, interference, layer 2 sleep capability, bandwidth allocation across time/frequency etc,. Given this, it is probably a little oversimplified to make a categorical conclusion.

Luke Thomas: Is it fair to say the IPR royalties for Mobile WiMAX is far lesser when compared to 3G LTE?

Eran Eshed: I don't think so. The IPR landscape in WiMAX is really unclear, and the policy by the WiMAX Forum is not very strict, so at this point it's a little difficult to predict what the royalty burden will look like when mass market happens. The bad experience the industry has from 3G is obviously the heavy CDMA burden imposed by Qualcomm, but I am not sure what is better – few parties with strong IPR positions or many with smaller IPR positions.

Luke Thomas: Will Mobile WiMAX merge with 3G LTE? Is there a demand for such a merger in the industry today?

Eran Eshed: In my opinion, it is unlikely that they will merge. LTE is really expected to be the more dominant and influential technology down the road, so it is more in the interest of WiMAX to potentially try and streamline with it. LTE, however, already adopted its own TDD version (being trialed in China nowadays), leaving WiMAX out of the 3GPP ecosystem. WiMAX, on the other hand, is now going FDD to try and capitalize on all of the IMT-2000 FDD spectrum that exists. Given this, it really seems that the political forces behind both standards are not very aligned with their priorities and strategic objectives.

Luke Thomas: What are the various applications/services that will be driven by next generation 3.9G technologies?

Eran Eshed: It is pretty difficult to predict which applications or services will become popular a few years down the road, I don't think that texting, IM or social networks like Facebook were sure bets at the time they were launched, and these things really have their own dynamics. What is clear today are the trends – the increase of user generated content, the growing popularity of social networking and various multimedia applications such as video or music. Overall, people are spending more “Internet time” on their mobile devices, rather than on their PCs. 3.9 or 4G technologies support this trend well with networks that offer increased throughputs and more symmetrical channels, higher capacity, and lower cost per bit.

Luke Thomas: Do you think users will be willing to pay a premium for next generation multimedia data applications?

Eran Eshed: I think that users will be willing to pay for services and applications they find valuable, so the real question is, how do carriers create value to the end user? In my opinion, wireless carriers should think of themselves more as wholesalers and less as service providers and open their networks to a much wider set of potential MVNOs and service providers, even non-traditional ones such as car manufacturers and broadcasters. This is the best way to spur the innovation and ensure that new services and applications are developed “outside the box.”

Luke Thomas: As Co-Founder and VP Marketing of Altair Semiconductor, what do you hope to accomplish in the next couple of years?

Eran Eshed: Altair is in the midst of a very important growth period. We launched our first mobile WiMAX product about a year ago and are now starting to book volume orders. In addition, we are trialing our XGP chipset in Japan and expect to start volume shipments in the fourth quarter of this year. On the LTE front, we will soon sample our first commercial ASIC which we expect to be one of the first of its kind in the market. We have expanded our global reach with offices in the U.S., Korea and soon in Japan and Taiwan. Clearly, there is a lot on our plate right now, but at the same time, we are laser focused on executing on the strategy we have outlined for the company. Over the next few years, we intend to turn the company profitable and capture the number one position in LTE semiconductors among startup companies.

Luke Thomas: How would you define success for yourself and for Altair Semiconductor?

Eran Eshed: Success for Altair would be to become the market leader in for all of its product lines. We already are technology leaders in WiMAX handheld devices, which is still a small market but is poised to grow significantly over the next few years. In addition, we are the lead supplier of XGP chipsets to Japanese service provider, WILLCOM, which for us is a captive market. The next big target is to sample our LTE chip this year, clear IOT with all the main base-station vendors and release the world's highest-performance and lowest power consuming LTE solution in the market. As the person in charge of marketing and business development at Altair, if we manage to achieve the above, we will be very satisfied.

Thanks for sharing your tacit knowledge and market insights Eran. It was a pleasure talking to you!