High-tech

Due Diligence: Artimi

By The Chilli analysts

Introduction

Artimi is a post-Chilli S2 fabless semiconductor company based in Cambridge, UK, initially targeting the ultra wideband (uwb, pronounced 'uweeba') market with a single-chip solution and reference design. The Chilli profiles Artimi with a view to assessing its potential for success, and makes a number of observations.

Artimi - vital statistics

Value proposition

Dynamics of the uwb market

The Chilli perspective

Artimi - vital statistics

Artimi was founded at the end of 2002 by GlobespanVirata alumni. David Baker, who is now chief operating officer (coo) and a hardware specialist, and Mark Moore, chief technology officer (cto) and software architect. They were later joined by Jack Lang, who became chief executive officer (ceo). Jack Lang has a track record as an entrepreneur, as founder of NetChannel (acquired by NTL), Electronic Share Information (acquired by E*Trade) and Perihelion Software. Richard Dellabarca, formerly an investment banker with ABN Amro, joined as commercial director.

During March 2003 the company raised S1/S2 funding of approximately £300K from close friends and family members. This has enabled the company to develop an initial demonstrator and prepare some patents.

The company is currently negotiating and finishing off the seed round. It is estimated that this will raise £500K+ of Chilli S3 funding. The company plans to raise Chilli R1 funding of circa £3M. This round will be used to expand the development team and deliver the first working solution and achieve some pilot customer traction.

Moore states that "Three quarters of the burn rate is salary", with Baker adding "We work hard to keep our costs low, and have acquired high-quality, high-spec second-hand lab equipment and used open source technology where possible."

Value proposition

Artimi plans to deliver a single-chip uwb solution. Uwb is a third generation of wireless technology that will supplement Wi-Fi and Bluetooth as a wireless personal area network (wpan).

So why does the world need yet another wireless technology when others are being deployed right now? The answer is that with uwb's high bandwidth over short distances (500mb/s under 5m, and 250mb/s under 10m), a new world of viable applications is opened up. There are still some open issues regarding interference and open standards, but we feel that this will be overcome by the time the company ships its first product.

So what does this high bandwidth allow you? It has the potential of solving that perennial last mile problem - distributing video around the house. It is purported that not only can you stream video around the house, but also you can have multiple channels being streamed. If the promoters of uwb can deliver their promise then some of those wires behind your television and set-top box (stb) may start disappearing. Currently, Wi-Fi solutions do not have the right amount of bandwidth, security or reliability to achieve this goal. Therefore uwb has great potential. Moore summarises Artimi's value proposition as "Broadly speaking, performance and range, from the both the receiver design/implementation and the built-in radio mesh networking support." What's a mesh network? A mesh network is an ad hoc, peer-to-peer network that continues to function even if some of its nodes die.

Moore declined to give a price indication, but research from The Chilli indicates a bom (bill of materials) cost for the chip and other components including the board, power supply and antenna in the $30-35 range for the initial phase. For phase two deployments The Chilli believes that there is a great potential for lower cost solutions that will take advantage of advances in process technology shrinks.

Uwb will be a worldwide market, requiring a parallel presence in all three theatres, namely America, Asia and Europe. Artimi aggressively plans to pursue all three in parallel, firstly by opening an office in the United States.

Dynamics of the uwb market

Studies are currently underway to see if uwb interferes with other systems such as Bluetooth, Wi-Fi (802.11a/b/g), umts and gps and vice-versa, but given that it uses very low transmit power relative to the other systems, it is unlikely that this will be a barrier to deployment, as uwb signals will appear as low threshold noise signals. Uwb operation is currently only licensed for use in the USA under FCC part 15 rules, with European legislation expected during 2004.

The standardisation of uwb for wpan is taking place in the forum of the IEEE. The proposals for the standard, 802.15.3a, will be voted on during July 2003, reducing the field to a handful of proposals, with a draft ready by the end of 2003. Artimi is an active participant in tracking the standards process and will have time to modify its design as required through a flexible design approach that will incorporate any changes.

The primary application of uwb is streaming video around the home, between consumer electronics and computing equipment, including stb and pvr (personal video recorder). Previous attempts at tackling this application failed because of:

• Cabling required around the house
• Multipath interference
• Lack of a content protection mechanism
• Adequate bandwidth and support for jitter-free video (quality of service)
• Cost per node

Uwb can overcome most of these inhibitors. Uwb does away with cables, has sufficient bandwidth, is not hindered by obstructions like walls, and can use the 802.15.3 qos protocol or adapt any of the high-level protocols originally developed for IEEE-1394 FireWire. To penetrate the consumer market, the cost per node must be low, with a bom under $10. Low power consumption will also be vital for battery-powered devices such as digital still cameras. The bandwidth, mesh networking capability and resistance to multipath interference will provide a lower overall system cost compared to Wi-Fi, which requires access points, and has a lower bandwidth. For an adequate home network, the distance between some nodes may be nearer 15-20 metres rather than 10m, requiring a slight variation of transmit power and bandwidth. One possible concern from content owners that will have to be addressed, perhaps under the WiMedia Alliance to promote wpan (similar to the Wi-Fi Alliance for Wi-Fi), is that while uwb signals propagate through walls, it does not mean that content paid for by one household will go through the walls to the next door neighbour, who has not paid for the right to watch that content. The attributes of uwb signals being hard to intercept and detect help, but it is a good idea to reuse whatever rights management system has been approved by the content owners (like the 5C scheme originally developed for FireWire), as approving a new scheme will take a long time.

Aside from video streaming in the consumer segment, another application will be in computing, for imaging and printing, where uwb will be a cable replacement for usb 2.0 and potentially FireWire. Other applications for uwb include radar and tracking which could provide some significant challenges for startups in this area, including Artimi.

Due to the compelling nature of uwb compared to Bluetooth and Wi-Fi, it has already attracted quite a few players. Each one with a unique design/product value proposition, including General Atomics, Staccato Communications, Time Domain, Wisair and Xtreme Spectrum. Essentially, Artimi will face three different sets of competitors:

• Large integrated device manufacturers (idms) including Philips, TI and STMicroelectronics. Artimi is likely to be an acquisition candidate, as several idms will have missed the market for uwb due to slow decision making and a focus on existing business
• Competition from existing startups with both a different background and design approach - Artimi is well-positioned to deal with this as they have management pedigree and a track record in delivering volume to the consumer segment, which most of the others don't have, although it would be advisable for Artimi to concentrate on some small niches in the beginning, until they have climbed the learning curve of delivering uwb
• A third set are startups that have managed to accumulate a large amount of funds. This is probably the tail end of the last bubble era when many vc (venture capital) firms were mesmerised by new technology and pumped in too much money without care for burn rate. As Artimi is fresh they are likely to have a lower burn rate and a leaner, meaner company culture than some of the bubble companies

The Chilli perspective

Two key questions to ask of Artimi are:

• Is there a long-term market for uwb that they can capitalise on?
• Is there any jam for them in the meantime?

Allied Business Intelligence (ABI) forecasts unit sales of 45.1m and revenues of $1.39bn by the end of 2007. Assuming that the IEEE standardisation effort continues at its current pace, standards-compliant chipsets will be reaching the market late 2004/early 2005, with end products coming out late 2005/early 2006, as bugs and interoperability issues are ironed out. ABI states that the consumer electronics industry will adopt proprietary uwb solutions prior to a standard being agreed; we at The Chilli believe that this is unlikely, as uwb, being a communications medium, will require interoperability with other uwb-enabled devices. The long-term market of video streaming in the house looks promising as uwb has the potential to remove many of the inhibitors (bandwidth, qos, cabling, digital rights management, multipath interference) that have dogged other schemes. Assuming a bom well below $10 is required and achieved by vendors then uwb will be adopted by both the high-volume consumer and computing segments in the same way as other interfaces have proliferated, e.g. irda, rs-232, Ethernet, usb, etc.

Given Intel's strategy of peripheral integration, and the number of US-based uwb startups, Artimi is well advised to not only focus on the large volume business but ensure that its intellectual property (ip) is ready for integration into other vendors chipsets. Artimi is unusual for a UK startup in that it is thinking about a global strategy from the beginning.

The consumer electronics vendors will appreciate the system-level design expertise and reference designs that Artimi can offer, for both uwb mesh networking, as well as coexisting with Wi-Fi, which some consumer vendors may be unfamiliar with.

Subcontract vendors in the Far East will be a key segment for Artimi, and attention to this area will set them apart from the 'Silicon Valley scramble' of startups saturating the Valley. In the short- to medium-term Artimi has an opportunity to get jam via system integrators and distributors.

What are the risks to Artimi? The issues of interference and coexistence, while important, are rapidly becoming a sideshow. Artimi has the pedigree to deliver low-cost devices to the consumer segment, so are likely to achieve that goal.

Longer-term, Artimi is a great candidate for merger and acquisition activity. This article has simplified the operation of uwb, but the technology is complex, and there are not many startups that can do this, setting a limit on the number of credible players in this space.

With a good pedigree, a proven track record, and experience of systems design (including digital, analogue and rf), plus the right financial/idm partner, Artimi has a good chance of providing some real returns to its original founders and initial investors. The team has clearly shown it has the ability to execute and deliver volume and low price.

Comments on this story? Send an e-mail to editor@thechilli.com

© Chilli Publishing Ltd 2003

14JUNE2003