ChipCrunch.com

At times it is easy to get wrapped up in following the latest technology trends, startups, funding and anything else related to emerging technologies in the semiconductor industry. Sure these things are interesting and often can be precursors as to where the industry might be heading, but just as important and interesting is it at times to delve into the details of bygone days, to dig up some interesting problems, solutions, and occasion some very entertaining history.

If assembly language is one of your favorite past-times, then you are likely to get a kick out of pagetable.com, a blog maintained by Michael Steil and Sebastian Biallas, related to assembly language, tricks, trivia and puzzles, and on some very interesting articles. Case in point is the latest entry about Bill Gates’ potential Personal Easter Eggs in 8 Bit BASIC.  This entry is not a simplye rumor, but rather a very well researched and detailed examination of Microsoft BASIC’s history on many different platforms, including Altair, 6502, Apple, Atari, and Commodore. Michael even throws in some assembly code that show how the Easter Egg was implemented. The article is a pleasure to read, full of history and nostalgia, and brings me back to my Atari and Commodore days - which are very much the reason why I ended up in the semiconductor industry in the first place. There are several other very interesting articles as well as puzzles that can be quite challenging if assembly is not your forte or if you are rusty like me.  Anyhow, if you have some spare time, abut don’t feel like wasting it mingling with your co-workers in the company 's break room, give this blog a look and indulge in some computer archeology.

If you are not actively involved in the telecommunications market you might be puzzled by the term femtocell. Sound like a bio-engineering term, or maybe a novel type of memory? Not even close. Essentially, it is nothing more than a small cellular base station which connects to a service provider’s network via a broadband internet connection. It is capable of simultaneously supporting a few mobile devices, and as such is mostly targeted at the small business and residential markets, where it can extend indoor service coverage. Femtocells are of quite some interest to mobile operators, who are always looking at extending the reach and capacity of their networks. Additionally, Forward Concepts is projecting a compound annual growth rate for femtocell based equipment of 126% over the next four years to $4.9 billion in 2012.

Where there are hot emerging markets and problems to be solved, you can always count on startups appearing to tackle them, such is the case with Percello. Founded in 2007, Percello is a fabless semiconductor company based out of Herzliya, Israel that specializes in digital baseband solutions for the 3G/HSDPA/HSUPA/HSPA+/LTE femtocell markets. Earlier this month, the company raised $12 million in Series B funding from Granite Ventures and Vertex Venture Capital which will be primarily used to finish the development of the PRC6000 Digital Baseband Processor for the UMTS femtocells.

The PRC6000, is Percello’s first chip in the Aquilo Family, which integrates all UMTS baseband functions on a single piece of silicon. The Aquilo architecture combines several programmable cores with dedicated accelerators and, according to the company, Aquilo chips can be cascaded to linearly increase the maximum number of concurrent users. As for the PRC6000, from previous press-release we know that Percello licensed the MIPS32 24Kc Core from MIPS Technologies in July. The 24Kc is a synthesizable core that runs at a clock speed of at least 700MHz in a 65nm process. Additionally, Percello also licensed the Ceva TeakLite III 32-bit DSP earlier this month, which runs above 550 MHz in a 65nm process. As such, it can be assumed that the PRC6000 is a combination of these two, in addition to whatever else Percello decides to integrate onto the die. It will support up to 8 simultaneous users and has a coverage radius of approximately 500m.  Mass production is expected to begin in the first half of 2009. Additional PRC6000 specification can be found here.

The company founders Shlomo Gadot, Yoav Volloch, and Rafy Carmon are all ex-LSI Logic, and beforehand were with Agere which was acquired by LSI in 2007.  As such they have quite a bit of experience with SOCs, nevertheless they will have their work cut out for them for as the market for femtocells takes off, many of the established players in the baseband market are likely to either adapt their solutions to this new segment or will go on acquisition sprees.

The American economy has definitely seen better days, no secrete there, but how to get out of this mess is another issue. Economic problems manifest themselves in many ways, with the most spectacular being the failures of giant corporations that are supposed to represent the bedrock of the economy. Well, if the giants are failing, why not encourage startups and entrepreneurship to get the economy going again? This is along the lines that Sramana Mitra seems to be thinking in her latest article at Forbes.com.

A quite successful entrepreneur in her own right, having founded and sold several software/web startups, and a strategy consultant in Silicon Valley since 1994, Sramana suggest that what the government really ought to do is give the entrepreneur a stimulus package. The package she is suggesting is an entrepreneurship-friendly tax policy targeted at entrepreneurs, angle investors, venture capital firms, and large corporation. In her article she examines the particular policies for each of the aforementioned groups in detail, but overall they can be summarized as follows: A tax-free pool of income for entrepreneurs and angel investors that can be reinvested into new ventures, an elimination of payroll taxes for small companies, a tired tax-structure for venture capital firms to once again encourages them to take some risks and invest in early stage startups in return for a lower tax on deals realized from those investments, and finally, tax breaks for large companies to invest into small business that in some form or other can in turn utilize the products and service offered by these large conglomerates. These are some interesting ideas indeed and definitely something to contemplate, especially when looking for more of a long-term plan for the country rather than a short-term stimulus package like the last one that was implemented by the goverment.

If you are interested in entrepreneurship in general you might also want to periodically take a glance at Sramana’s own blog in which she covers a vast spectrum of Silicon Valley from web startups, to entrepreneurship strategy and stock discussions, and even the occasional semiconductor blurb. As a matter of fact, her post in mid June which analyzed Cadence’s hostile takeover attempt of fellow EDA tool develop Mentor Graphics caused quite a discussion amongst her readers.

Venture capital money has been tight for semiconductor startups recently, but as always there are exceptions to the rule. One hot area is digital amplifiers. We wrote about digital amplifiers and BlackSand technologies in particular here and here. The other hot area is programmable transceivers, which is where BitWave Semiconductor comes into play. The company was formed in 2004 with the mission to assist cell phone manufacturers by reducing the number of chips needed to support all the different wireless standards, thus enabling them to offer more features in a smaller package and at a lower price. If you don’t think there are that many wireless standards, you might want to reconsider: GSM, WCDMA, HSDPA, HSUPA, AMPS, NAMPS, CDMA, CDMA2000, 1xRTT, EV-DO, CDPD, DVB-H, MediaFLO, DMB, Wi-Fi, WiMax, WiBro, UWB, Bluetooth, GPS, DAB, AM/FM, DECT, etc.

The solution to this mess, at least according to BitWave, is what the company calls Softranceiver Technology, which according to their website is the combination of “the art of analog with the science of digital.” Essentially, the Softranceiver architecture integrates digital control for many of the analog components. Thus, an API can be utilized to set proper registers which in turn control the output of the analog blocks. The result is that a single transceiver can be programmed to support many different wireless standards, significantly reducing the component integration costs. BitWave’s approach differs significantly from software-defined radio (SDR) in that it does not rely on digitizing the incoming signal via an analog-to-digital converter (ADC) and then processing it via a DSP, which according to BitWave’s President and CEO Dr. Michael Farese, leads to high-power consumption and has not been implemented successfully in a mobile device. An interview with Michael in which he describes several additional Softranceiver advantages can be found over at WTRS.

Currently, the Softranceiver is slated for volume production and general availability in the second quarter of 2008, and will be branded as the BW1102 Softranceiver RFIC. To make the commercialization happen, the company raised an additional $10 million in B-round funding from Apex Venture Partners, TVM Capital, and Ecentury Capital Partners last month. Clearly, the applications of this technology extend far beyond cell phones to just about any device that requires multi-mode, multi-band wireless capability. The potential market for this technology seems very appealing given the current growth rate for mobile devices. As long as the API is reasonable, and the transceiver works as promised, the potential rewards for the company could be enormous.