Lessons Learned from Transmeta’s Death and Rebirth

Transmeta makes for an interesting Silicon Valley case study. The net of its announcement last week is that they will continue to be with us indefinably.

This company has an amazing history, and its story demonstrates the opportunities and mistakes that can surround a high-profile startup in Silicon Valley.

The Alien Influence: Marketing at Its Best

When Transmeta first appeared on the scene, it was housed in a building across from where I was working at the time with its windows blacked out, cloaked in massive secrecy. Rumors included — and I’m not kidding — that they were developing technology that had been derived from alien spaceships that would change the world forever. Segway later used a very similar process to drive interest in their offering to incredibly high levels, and both efforts set the benchmark for low- to zero-dollar marketing resulting in massive media coverage.

The underlying concept taps into the human tendency to want what is unavailable — in this case, information. If you can make something attractive and then dangle it just out of reach of the audience, you can garner an almost rabid response. This creates a virtual media feeding frenzy as reporters fight to get the “scoop” on the breaking news. Contrast this with what normally happens with a new release where the overworked reporter, after seeing the 30th “unique product” in a given week, has a hard time staying awake during the briefing, let alone writing about that product favorably — or at all.

As the announcement of Transmeta’s product drew close, the major news services were fighting each other to get the story out. To this day, I’ve never seen the kind of coverage Transmeta had at their launch. Had it been done right, Intel probably never would have been able to respond, and Transmeta could have had a near impregnable beachhead in the laptop market they could have used to challenge Intel broadly.

In War and Marketing, Timing is Everything

Unfortunately Transmeta fired too early. If they had just had one branded customer on stage with them during the launch that was announcing a product, the news stories would have been all about the massive damage they would do to Intel and their success. But Transmeta didn’t do this, and the stories spun south and instead focused on an apparent lack of customer support. The product that actually had been shown on stage never, to my recollection, actually showed up.

The real damage was that the presentation did scare Intel half to death, and subsequently, they put a massive amount of resources into creating what eventually became the Pentium M. This was a ground-up, redesigned product that, while not as power-efficient as Transmeta’s, hit the sweet spot of the laptop market well enough and removed much of the potential that Transmeta otherwise would have had.

Strangely enough, consumers actually saw a much better benefit then they likely otherwise would have as the Intel part moved more rapidly into a number of products and average efficiency and battery life jumped dramatically. However, Intel also got much of the financial benefit that otherwise would have been Transmeta’s.

Rebuilding the Wave

After this, Transmeta slowly built up more and more vendor interest, capturing Sony, Toshiba, and HP in what were market-leading products in various segments. Their first offering, the Crusoe, was adequate for Windows 2000 and Windows XP, provided vastly better thermal and battery performance then anything from either AMD or Intel, and was used in very high-profile products from the branded vendors mentioned above.

Step by step, Transmeta was rebuilding its image, now the hard way, and once again it was looking like a player. At the time, Intel was having trouble getting its new mobile processor out the door and its “Speed Step” technology (basically a stop-gap measure that cut processor performance dramatically in order to get barely acceptable battery life) was seen by many as a bad joke.

Meanwhile, IBM and others were rumored to be moving to Transmeta, and Transmeta was planning to bring a new chip to market, called Efficeon, that was vastly more powerful then Crusoe and capable of turning the company into a real player again.

Shot from an Enexpected Place

Transmeta had been using IBM to fabricate its parts, but like others who have looked to IBM for this, it was having recurring cost and delivery problems even though quality was first rate. In addition, many think that using IBM Microelectronics is the path into IBM hardware and this has, time and time again, has proven to be untrue.

Amid increased buzz in its favor, the firm moved to go public. News surrounding Transmeta was very positive, and the possibility that IBM would be the next company to use its technology was fueling a potential buying frenzy. The result, had it gone well, was expected to be a massive infusion of cash to a degree that Transmeta could have purchased AMD and the combined entity would have given Intel an instant heart attack.

At the peak, and right before the IPO, IBM made the surprise announcement (I’ve never known IBM to do this) that they would not be using Transmeta. While never proven, this was alleged to have resulted from a panicked request from Intel, which if true, was incredibly well timed to pull the rug out from under Transmeta. This, in turn, meant they generated a fraction of the capital they had anticipated which made it impossible for them to acquire any larger firm, let alone AMD.

Transmeta Stumbles Again

Before bringing out its new high performance processor and with companies like Toshiba basing product releases on that part, Transmeta stopped using IBM to fabricate its chips and found a lower-cost operation in Asia. Unfortunately the transition did not go well and Efficeon did not make its critical fourth-quarter launch date to the embarrassment of the PC product managers who depended on the part.

Additional commitments dried up, and the Toshiba relationship went into rapid decline. Transmeta was once again on the ropes. Still, once the part finally arrived, it performed as advertised and several Sharp branded products we’ve tested using it have been very impressive. However, at the end of 2004, Transmeta was clearly in trouble — costs were significantly exceeding revenues and new opportunities like the amazing OQO pocket laptop were not making up the difference.

Transition and Opportunity

The market continues to favor power conservation as a long-term goal, not only in portable devices but in products that traditionally were thought not to have this requirement. This is because current generation mainstream parts have become so powerful that the heat they generate and the power they utilize have become problems in markets ranging from consumer to corporate, from full-on PCs to embedded appliances. Transmeta just needs time for players in these markets to understand the timely benefits these parts provide, as well as time for them to regain their trust in the vendor so they can return it to growth.

The dramatic changes Transmeta just underwent, which cut their ongoing costs and provided sustained revenue sources, provides them with this much-needed time. The company’s licensing efforts should better allow its technology to proliferate — and there is a very good chance we will still be talking about Transmeta at the end of the decade as a result.

Still, it is interesting to look back at what might have been and remember that timing, the right partners, and execution remain the only sure way to success in this, and any, market.