Live blogging Renesas DevCon

Wed., 3 p.m.: Here comes a new low power RX version.Much of the chest-thumping at Renesas DevCon (see Patrick Mannion's excellent overview) this week was about the relatively new RX microcontroller family, a hybrid RISC-CISC architecture which was rolled out earlier this year. But expect a major rev come Q1 2001 in the form of the RX200, which will aggressively turn the crank on low power.

The RX200 will be a 32-bit, 50MHz MCU with the same CPU core as the RX but optimized to reduce power draw and enhanced with new analog functionality. It will include up to 1MB flash and operate at 200uA/MHz and all the way down 1uA in standby mode.

How? Having access to the 78K family and the VH850 family after the NEC deal has been interesting, one company source said. The 78K, for example, claims 190uA/MHz and 0.3 uA in standby mode. The higher-performance v850 claims 600uA/MHz and 1.5uA in standby.

This is great stuff but something the company isn’t particularly keen on talking about, even though plenty of online material touts portions of the story and Shigeo Mizugaki, senior vice president who runs the MCU business unit, mentioned it prominently at lunch. Stay tuned for more as we toss out our 2010 calendars in a few months.

Wed., 9:45 a.m.: Ritesh Tyagi talks about non-volatile memory roadmaps. Flash and other NVM is key to Renesas' MCU strategy because no matter how much you crank the clock, you can leave performance on the table with slow memories.

In the coming years:

• Renesas will push its performance MONOS flash, which is now running at 100 MHz on a 90nm process, to 120MHz on a 40nm process by 2013 and 150MHz on 28nm by 2015.
• The company will launch MRAM capability, so-called "unified memory" where code and data can be managed by a single memory (instead of using SRAM and flash).  MRAM is expected to come out at 150 MHz on a 90nm process in 2013, moving to 200 MHz on a 40nm process in 2015.

Tues., 9:15-ish p.m.: Designers by day, rock stars by night.



Tues., 3:30: Renesas’ power products business will launch 1,000 products in the next three years as part of an aggressive campaign to drive double-digit revenue growth.

“Five hundred will be low-voltage mosfets, 250 will be IGBTs, and the rest will be low-voltage and high-voltage triac devices,” said Tad Keeley, senior director in Renesas’ power business. “The second component of that growth strategy is we’re going to build our global structure outside Japan, focus on China, but not just there.”

Why?
In a keynote yesterday, Daniel Mahoney, president and CEO of Renesas Electronics America Inc., said, "We think that every microcontroller customer is a potential MOSFET customer."

Power devices is one of five groups within the Analog and Power business group. Power devices make up 25% of A&P, and A&P contributes 31% of Renesas top line. Power devices will contribute $800-$900 million of the $10 billion Renesas revenue stream, Keeley said.

At first glance, the growth goal is no biggie. Growth in power from 2006-09 was on the order of 12 %, but Keeley pointed out that they want to see significant double-digit growth in their area, not just 10-11 percent per year. He and Tetsuo Sato, director, business development for standard products, are bullish about the growth prospects across all power’s technology lines: low-voltage MOSFETs, IGBTs and triacs and high-voltage MOSFETS. He sees growth rates of at least 10 percent per technology line.  They want to strengthen sales outside of Japan to 54% outside of Japan, 20 percent alone in China.

Tues., 2:15: MUCH wailing, gnashing of teeth, foot stamping echoes throughout hotel lobby. What the heck? Thought someone didn't get a cool Renesas tchotchke on the demo floor, but it turns out we're just a few blocks from Disneyland. Many future engineers more interested in seeing Mickey than microcontrollers, it seems.

Tues., 12:45 p.m.: Omid Milani talks as fast as the possibilities that pop into his mind on the floor the Renesas DevCon. As the director of Renesas' display business unit, he ponders the application possibilities as he shows off two 3D displays, one a 3.1-inch panel for the cell phone market, the other a 7-inch display that has potential in medical, industrial and consumer applications.

"Video on a handset is obvious," he says, "where you could watch 'Avatar' while you're sitting on a bus." After that, the possibilities become conjecture. They include:

• GPS applications in which your destination "pops out" from the map behind it
• Giving the user the ability on a touch screen to dive into documents or applications rather than having to scroll through them.

There will be others, of course, that we can't envision yet. Renesas and NEC have gotten here by employing a technology it calls Horizontally Double Density Pixel (HDDP). In short, it doubles the amount of horizontal pixels and then divides each pixel into two sub-pixels. The system then writes either a 2D or 3D signal to each subpixel. This among other things allows both 2D and 3D images to be displayed on the same screen. (Traditional LCD modules require two pixels to display a 3D image, cutting resolution in half).

Milani claims the Renesas/NEC approach doesn't appreciably worsen power dissipation.



This Blackberry camera photo of the 7-inch display doesn't do the 3D justice obviously but the technology does work and it looks very cool. That said, it remains to be seen not only what type of useful applications will emerge for 3D displays and how long the human eye can stand to look at it.

Milani says one of the beauties of the HDDP technology is that because it renders both 2D and 3D, you can use the screen in traditional 2D mode whenever you want without losing resolution.