Antony M. Fadell (born 1969)\

Image: Tony_Fadell

From February 2001, he was responsible for the development of the iPod as part of the Special Projects Group and iPod under the leadership of Jon Rubinstein. In 2004, he became vice president of iPod, and in March 2006, he replaced Rubinstein as senior vice president of the iPod division, who retired from Apple. In November 2008, Fadell left Apple and in 2010 founded Nest, a company developing a self-learning thermostat for heating, which Honeywell sued that same year for infringing on its smart thermostat patents…

Fadell secured a hardware solution for the iPod when he agreed with the founders of a small company called PortalPlayer that their industrial design would be used for the iPod. He had previously collaborated with PortalPlayer and was familiar with their solutions. He also knew that it would be possible to create a player the size of a “cigarette pack” based on their design, which was the current requirement for the iPod. Most importantly, PortalPlayer had experience designing hard drive-based players and knew how to deal with the pitfalls of buffers and player stuttering.

This was precisely the problem that plagued Apple. They encountered portable audio CD players, let alone portable CD MP3 players, and for a long time, the issue was addressed with varying degrees of success by storing pre-loaded data in a buffer memory. When designing the PJB-100, the Compaq team was able to use 12MB of DRAM memory, which is erased in the event of a power failure, but in this case, that doesn’t matter. It is cheaper, faster, and less energy-intensive, which is important. Such a large memory allows two songs to be stored, so the device only needs to guess which one will be played next in the sequence. The disk can then remain idle for almost two entire songs, meaning not only is it free from vibration, but it also consumes no power.

Apple and PortalPlayer decided to use 32 MB SDRAM for the iPod, which would allow it to protect against outages for about five songs, or approximately 20 minutes of playback, and would also allow it to turn off the hard drive and save energy. At the same time, Apple developed a system to predict what the user might want to play based on which songs they skipped, so that the most likely songs the user would want to hear were loaded into the cache.

Under the agreement with PortalPlayer, Apple had immediate access to the hardware design, including the software. Several prototypes existed, and work on the reference design was about 80% complete, but there were still some issues and things to be finished. For example, it was not possible to create playlists longer than ten items. In addition, both the appearance of the reference device and the software looked like software produced by programmers and engineers, not something that Apple designers would create.

To Fadell’s credit, he persuaded PortalPlayer to work exclusively with Apple. At the time, the company had more than ten customers for its reference design, which was used by a number of Asian companies. However, PortalPlayer also had a request from IBM for an MP3 player that would use Microdrive, a tiny one-inch hard drive developed by IBM, and even offer wireless headphones. Paradoxically, Apple began using Microdrive drives in 2004 for its new iPod Mini, but by then the drives were being manufactured by Hitachi, which IBM had sold the production to. All 280 PortalPlayer employees (many of whom worked part-time, with 80 people employed by the company in India) began to focus intensively on finalizing the hardware design for the iPod and its firmware. It was necessary to include not only compatibility with the FireWire interface, but also the AAC and Audible formats for audiobooks, and above all to redesign the interface.

All that remained was to find an operating system for the player, a higher layer placed above the player’s firmware itself. Here, too, Apple turned to a company founded by someone associated with Apple who already had the necessary product ready. That company was Pixo, co-founded in 1994 by Paul Mercer, a former developer of Finder for Mac OS 7. The company focused on developing a platform for wireless devices, primarily mobile phones. It had a development environment for developers and also the Pixo OS platform, which served as an integrated operating system for mobile devices, offering a directory and internet access. At that time, its main customer was Samsung, to whose CDMA phones it supplied a built-in operating system, framework, and development environment. And it was precisely this kind of comprehensive system, enabling both rapid application design and development and the built-in operating system itself, that Apple needed. So it reached an agreement with Pixo. How this was done remains a mystery to this day. The first iPods still mentioned the licensing of Pixo OS, but this reference disappeared in later models. The most likely scenario is that Apple bought Pixo OS outright, probably along with part of the development team. The fact remains that with rapid prototyping and testing of application appearance and menu design using Pixo OS, Apple managed to meet the tight deadline for launching the iPod.

The context of this story

Antony M. Fadell (born 1969)\

Table of contents