Sep 20, 2003

Motorola Semi's Failure to Innovate

Motorola struggling isn't news to anyone who watches Apple, nor is it news to anyone else with a vested interested in Motorola's semiconductor branch. But it's not like the company is beleagured. It's not like it takes Motorola's last gasping breath to release a new G4. The problem with Motorola's struggle is Motorola's attitude toward innovation.

Since the joint IBM/Motorola Somerset facility was turned over to Motorola, all innovation within its PowerPC division has stopped. "Don't mess with a good design," you say? Consider this: the last new cores Motorola helped to create were the the G2 contingent: the PowerPC 602, 603, 604, and 620 chips and their variants. Working with IBM on their last joint project, Motorola produced the 750, a tweak to the 603 core. Since then, every new chip has been based on something that came from that effort.

The entire G4 family is naught but a 750 core with a SIMD unit and an FPU ripped from the PowerPC 604. While not a bad chip, it's obvious that there was something inherently stifling about the G4's architecture. Remember the 500 MHz Fiasco? It's easy to understand if you look at the 74xx series as a 603 with major kludges thrown on top. That they even reached past 1 GHz is impressive.

Motorola's new PowerQUICC III family (rumored to be the PowerPC G5 for some time) is also based on the 603 core, as were the PowerQUICC I and PowerQUICC II families. Zooming out for a broader perspective, PowerQUICC III and Motorola's automotive PowerPC 52xx line are both based on the e500 core, more or less a 603 core with tweaks to meet the BookE spec.

The above examples are not just symptoms, however. Motorola innovates rarely and rides the wave for as long as possible, even when it's clear that they need to get back to the drawing board. They've been extending 1994 technology for ten years, and while that may work in some cases (IBM still successfully updates its own PowerPC 750 variants), in the end it serves only to decrease the quality of its product. Motorola is shooting not just itself, but also its customers, in the foot.

In the end, Motorola's upper management is at fault for making the decisions not to innovate; the engineers and designers at Motorola do an excellent job of what they have to work with. Hopefully with CEO Chris Galvin's resignation the attitudes and practices of management will change and Motorola can produce some astounding new cores for the PowerPC family.

No comments:

Post a Comment