Turbulent Start for Transmeta
Since emerging into the limelight with a widely covered public announcement [see “Linley on Linux” April 2000], Transmeta has had a busy year. The company began shipping its first products, gained several significant design wins and successfully executed an IPO. Although its stock price has since been cut in half, Transmeta still carries an impressive market cap of about $3 billion US. The company has seen tougher-than-expected competition from Intel in the PC market as well as slow movement in the webpad market. On the other hand, Transmeta recently opened up a third front in the server market. The startup, also famous for being Linus Torvalds' employer, produces the Crusoe chip that is compatible with Intel's processors but uses less power. This accomplishment is due to its unique code-morphing technology that emulates Intel's architecture on simpler VLIW hardware. Initially, Transmeta believed that Intel could not match Crusoe's low-power consumption without developing its own code-morphing technology, which would take years. But Intel fought back quickly.
Without access to code morphing, Intel instead took advantage of its superior manufacturing technology to reduce the power consumption of its chips. Last June, Intel rolled out new processors that operate at lower voltages than Transmeta's IBM-built chips (IBM has manufacturing technology similar to Intel's but reserves it for internal products). Although some studies show that Crusoe still has an edge in power savings, its advantage over Intel's latest chips is small.
Transmeta has also failed to produce any useful performance data for its processor. To be fair, code morphing makes benchmarking difficult; the Crusoe chip actually gets faster if you run a program more often. Testing of initial Crusoe systems found their performance to be roughly that of a mere 300MHz Pentium III, but these tests ran the benchmark program only once. How fast the chip would run with more “practice” remains unclear, as Transmeta itself still isn't talking.
With its core messages of performance and low power under attack, Transmeta made limited headway in the notebook PC market over the past year. The company placed Crusoe in new ultralight PCs from Sony, Fujitsu, Hitachi and NEC. These systems are aimed mainly at the Japanese market, where space and power savings are at a premium. The leading makers of notebook PCs for the US market—IBM, Toshiba, Compaq and Dell—have yet to roll out any Crusoe-based systems, although all have evaluated Transmeta's technology. As a result, Transmeta's share of the ultralight PC market is perhaps 10% so far. Since that market is just a small fraction of the PC market, Transmeta holds far less than 1% of the total PC processor market. Thus, the company poses little threat to Intel's revenues. Transmeta doesn't need much of Intel's market; however, about 50% of the ultralight PC market would help Transmeta break even.
Investors are hoping for more than breaking even, as Transmeta has lost more than $200 million since it was founded and will probably lose another $50 million, even if it reaches its goal of profitability by the end of this year. Investors want some payback. That's where new markets come in. Transmeta's second target market is webpads. These portable devices access web content through a wireless link, showing it on a flat-panel display. Webpads typically use PC-compatible processors, allowing them to run all the plug-ins on the Web but don't need to run Windows; many use Linux instead. The good news is that virtually every webpad announced to date uses Crusoe for its unique combination of low-power consumption, PC compatibility and low price. Companies such as Gateway, Phillips, Hitachi and Frontpath (formerly S3) have selected Crusoe for their webpads.
The bad news: webpad sales have been minimal to date, slowed by prices that are often $1,000 or more. The most expensive part of these webpads is the flat-panel display, which can cost several hundred dollars alone. However, the recent slowdown in PC sales has reduced demand for flat panels, causing prices to drop. With panel-manufacturing capacity continuing to increase, these prices should fall dramatically and help webpads become more popular in 2001. Still, Transmeta expects webpads to contribute less than 20% of its revenue this year.
Transmeta recently added web servers as a third target market. Although Crusoe is not as powerful as server processors from Intel and others, the low-power chip doesn't need a bulky heat sink or other cooling equipment. This enables vendors to pack several Crusoe chips into the space required by a single power-hungry Intel chip. Since the task of serving web pages can be easily divided among several processors, Crusoe makes up in numbers what it lacks in individual performance.
So far, only a single small company has endorsed Transmeta's server vision, and Transmeta expects little server revenue in 2001. But this rapidly growing market represents another opportunity for Crusoe, even if Intel can block Transmeta in the PC market. Transmeta hopes PC revenues can tide it over until these new markets begin to flourish. In 2002, we may finally see Transmeta's full business plan in action.
|Non-Linux FOSS: libnotify, OS X Style||Jun 18, 2013|
|Containers—Not Virtual Machines—Are the Future Cloud||Jun 17, 2013|
|Lock-Free Multi-Producer Multi-Consumer Queue on Ring Buffer||Jun 12, 2013|
|Weechat, Irssi's Little Brother||Jun 11, 2013|
|One Tail Just Isn't Enough||Jun 07, 2013|
|Introduction to MapReduce with Hadoop on Linux||Jun 05, 2013|
- Containers—Not Virtual Machines—Are the Future Cloud
- Non-Linux FOSS: libnotify, OS X Style
- New Products
- Validate an E-Mail Address with PHP, the Right Way
- RSS Feeds
- Introduction to MapReduce with Hadoop on Linux
- Lock-Free Multi-Producer Multi-Consumer Queue on Ring Buffer
- Help with Designing or Debugging CORBA Applications
- Returning Values from Bash Functions
- Linux Systems Administrator
- Welcome to 1998
5 min 49 sec ago
- notifier shortcomings
29 min 31 sec ago
2 hours 6 min ago
- Android User
2 hours 8 min ago
- Reply to comment | Linux Journal
4 hours 1 min ago
6 hours 50 min ago
- This is a good post. This
12 hours 3 min ago
- Great, This is really amazing
12 hours 5 min ago
- These posts are really good
12 hours 7 min ago
- It’s a really great site you
12 hours 9 min ago
Free Webinar: Hadoop
How to Build an Optimal Hadoop Cluster to Store and Maintain Unlimited Amounts of Data Using Microservers
Realizing the promise of Apache® Hadoop® requires the effective deployment of compute, memory, storage and networking to achieve optimal results. With its flexibility and multitude of options, it is easy to over or under provision the server infrastructure, resulting in poor performance and high TCO. Join us for an in depth, technical discussion with industry experts from leading Hadoop and server companies who will provide insights into the key considerations for designing and deploying an optimal Hadoop cluster.
Some of key questions to be discussed are:
- What is the “typical” Hadoop cluster and what should be installed on the different machine types?
- Why should you consider the typical workload patterns when making your hardware decisions?
- Are all microservers created equal for Hadoop deployments?
- How do I plan for expansion if I require more compute, memory, storage or networking?