Some of those customers may be attracted by the new integration
strategy that's been backed up by smart investments. Before 2007,
Maxim had done almost no acquisitions, save for acquiring a
Tektronix wafer fab. It preferred to build everything in house.
Since 2007, it's acquired a dozen companies, including
AG for $130 million plus $34 million in debt.
SensorDynamics focuses on low-power wireless interfaces to MEMS
sensors and is best known for inertial MEMS sensors and
automotive smart-key chips.
which makes a 65-nm CMOS transceiver IC for HSPA and LTE
which makes transceivers and transimpedance amplifiers span the
entire spectrum of data rates, from 1- through to 10-Gbps.
a small firm that was developing and licensing digital
control technology based on ''state-space."
What lies ahead
It's a well-framed story, and the actions Doluca and his team have
taken since 2007 make logical sense. But challenges abound,
especially because pursuing an component-integration strategy
requires a completely different mindset about engineering, according
to Steve Ohr, analog analyst at Gartner.
"They need to figure out the best balance between standard analog
products (multi-market building blocks) and application-specific
analog (like cell phone power management ICs or smart meters). The
product lines require totally different engineering processes, and
offer quite different revenue streams and margins," he said.
Traditional analog parts are usually designed by a single engineer,
require little promotional outlay and can be sold for decades. The
gross margins can be as high as 70 percent, Ohr said.
"With application-specific parts, you’re talking about a
specialized, more integrated part type, put together with a short
time-to-market window, and often for a specific model consumer
product with a 12- or 18-month life cycle," Ohr said.
"Engineering-wise, this is a coordinated team effort. But an
ASIC/ASSP purchaser like Apple will know what your cost structure
is, and will beat you up on pricing."
A company may reap hundreds of millions in revenue on an ASIC/ASSP
product but with little margin, he added.
"The reality is that very few companies are equipped to do both
types of analog--standard and custom application-specific," Ohr
There are old maxims about company transformation: One is "time will
tell." Another is "Don't count your chickens before they hatch."
"Never base your design on parts not in production. You're asking for trouble. If you need prototype parts to tweak your design, you're asking for trouble. Never promise more than you can deliver."
Maxim's problem has always been that they have a lot of devices listed as "in production," yet you can't actually order them in anything but huge quantities.
This is why we always get purchasing to confirm availability of anything before it gets designed in, but one can also use the old rule of thumb: "don't design with it if you can't buy it from DigiKey or Mouser."
The problem I have always had with Maxim is getting any production parts at all. Parts were always available for prototyping but when it came time for production the parts were never available. Designing a Maxim part into a design was the equivalent to cutting you own throat.
"Before 2007, Maxim had done almost no acquisitions, save for acquiring a Tektronix wafer fab. It preferred to build everything in house. "
What about Dallas Semiconductor in 2001 for $2.5 Billion in stock?
"Maxim shares have begun to outperform some key competitors." I dont think that is the case.
Atleast a cursory look at yahoo finance comparing Maxim, LTC and ADI for the past year it shows equal or slightly weaker performance of the Maxim stock...thus depending on when you start the comparison the results will change :)
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.