Archive for September, 2009
Saturday, September 19th, 2009
Headquartered in Germany, X-FAB is a foundry with manufacturing operations in Erfurt and Dresden, Plymouth in the U.K., Lubbock in Texas, and Kuching, Sarawak in Malaysia. On September 8th, X-FAB made an interesting announcement with the introduction of their XA035 analog/mixed-signal CMOS high-temperature process for ICs that live tough lives: devices that operate in ambient temperatures of up to 175 degrees Celsius in places like the underside of your car.
I had a chance to speak this week by phone about the XA035 process with X-FAB Vice President of Business Development Mark Miller. Miller has a long and distinguished career in EDA with a track record that includes Business Development VP for DFM products at Cadence, VP at TeraSystems and at Synchronicity, and director of the IC Tech Center for Mentor Graphics. These days, Miller’s world is about manufacturing and that’s where our conversation started:
Mark said, “If you look around anyone’s living room, you’ll see a lot of electronics — high-definition TVs, computers, gaming stations, and so on. Our XA035 process, however, is targeted at something different. Of course, the process can be used for ‘standard’ ICs, and for non-volatile memory, [but it also addresses] a wide voltage range that extends beyond the usual digital foundry [offering].
“XA035 is focused on what we call the ‘hidden’ semiconductor markets. Extremely important, these markets include automotive, avionics, and military applications, all of which have as a key attribute the ability to meet stringent qualifications. Interestingly, it’s automotive reliability that involves a lot of such hot and harsh environments, even more so [than some other markets].
“If you’re building a control system in a car, for instance, the electronics have to withstand the underside of a car, where temperatures can go as high as 175 to 185 degrees C. That’s [also the case with] your braking system, another application area where reliability is critical. Just think of a designer being able to use a process technology like XA035, one that meets [industry standard specifications such as] AEC-Q100].
“But, there’s even more in our announcement. If you’re a designer, you also want to be able to analyze your design inside of its [target] environment, along with temperature-qualified models. You want to be able to cycle things 10,000 times or more, to look for dielectric breakdown and stress migration. That requires a lot of extended modeling. Yes, it can be expensive, but if you’re a designer for chips under the hood, you have no choice. Today’s [automotive standards] are not just a list of features, they also include a set of tests and reliability requirement. Most major auto manufacturer [understand this], especially in developing mission critical features.
“It’s outside of [the usual foundry business model] to offer analysis tools, but we have put together a reliability simulator and modeling tool that will take your particular design, establish a mission profile, and then go through and model it for you. The tool [allows you] to understand what the lifetime reliability parameters are, which is a really interesting and novel approach.
“Some products degrade in a step-function failure mode, while others may [exhibit] a gradual decline over a 20-year period. The designers can see all of that with our tools, and therefore understand how to tweak the design to meet the parametric degradation [predicted by the simulation].
“High reliability electronics is a very special applications area. Something a designer develops might be used [in a system] for the next 20 years. As a result, [we see] the designer’s paradox. If you have the responsibility for designing such a product, how do you model it? How many times will it be used. What is the mission profile? With our tools, the designer can take the specifications, and all the models, to get a very clear picture of the reliability. The tool interfaces directly to the device specification and characterizing — it’s pretty unique.”
I asked Mark if analog and mixed-signal devices are more susceptible than digital devices to degradation when deployed in harsh environment.
He said, “It turns out that [in general], the variety of devices manufactured in an analog/mixed-signal foundry is substantially different from what you make in a standard digital-type foundry. We’re not just talking about 3 to 5 volts here. Some of these analog/mixed-signal products [endure] up to 500 or 600 volts. Extensive modeling of those devices [during development] is critical.
“Analog and mixed signal have always been a great deal more work than standard digital. It’s the analog/mixed-signal product that has to interface with the real world. That kind of development requires pretty rigorous engineering and process development.”
I asked Mark if X-FAB headquartered in Germany because of the automotive industry centered there. Mark, who works out of Silicon Valley, said, “Our offering, and the XA035 process in particular, is not specifically an automotive application. In addition to Germany, we also have manufacturing in Malaysia and Texas. X-FAB is very much a global presence, but a lot of the history of the company is centered in Germany.
“There are other independent foundries in the world, but we’ve been in the business for longer than anybody else in analog, mixed-signal [domain]. Working in the general area of high reliability is not new for us.
“The XA035 [really speaks to all of that]: the overall high reliability of devices built on the process, the wide variety of device sizes available including non-volatile memory, and the tools for degradation modeling and developing a mission profile.”
Mark concluded, “There’s a lot of artistry, and creative, innovative design that [characterizes] the analog domain. At X-FAB, we feel that we are very much in that tradition and contributing to its growth.”
The Press Release …
X-FAB Silicon Foundries announced XA035, an analog/mixed-signal CMOS high-temperature foundry process for ICs requiring operating temperatures up to 175 degrees Celsius. The modular 0.35 micrometer process is the first to enable high-temperature-capable SoCs, combining high-voltage and embedded non-volatile memory elements.
Design support includes the XA035 Lifetime Calculator, a tool that calculates expected IC lifetime for a given mission profile to help determine lifetime/temperature trade-offs.
The new XA035 high-temperature process surpasses the stringent AEC-Q100 qualification tests for automotive IC quality and reliability, and is ideally suited for high-precision analog circuits, sensor front-ends, and brushless DC motor controls for automotive, industrial, aerospace and military markets.
X-FAB offers XA035 design kits covering all major EDA platforms. It also features a variety of dense standard cell libraries optimized for area, speed, low power or low noise; and I/O libraries, including ESD support. All libraries take the temperature effects into account. In addition, XA035 supports parasitic diode modeling. This new feature enables pre-layout parasitic diode leakage simulation and lets designers simulate leakages at high temperatures early in the design flow.
Tuesday, September 8th, 2009
Tuscany Design Automation, based in Colorado, has been around for over 4 years, but is only now concentrating on making itself known. Part of that process involved bringing EDA veteran Keith Mueller onboard as CEO. Given his long tenure in the industry, Mueller knows full well what it takes to establish a message and the right business model for a successful EDA startup, and by the sounds of things he relishes bringing that knowledge to Tuscany DA.
I spoke by phone with Mueller recently about his background and his new corporate home. He told me that launching a new image and logo for Tuscany was somewhat of a logistical challenge in the 4 short weeks between his arriving at the company in mid-June and DAC 2009 in San Francisco in late July.
“Nonetheless, we pulled it off and had a great response at DAC!” he said.
Mueller added, “Since its inception, the people at Tuscany have been focused on working with early customers to develop tools for structured design and perfecting their core infrastructure and database – focused on the technology, not the messaging. They had never done a press release prior to my joining the company.
“For DAC, we decided to officially launch the structured design tool called Tego, as well as the Tuscany Dashboard, a product with a much broader market built upon their core infrastructure. Tuscany is now definitely on the radar within the industry.
“We’re going to have a broad appeal to those design houses who manage remotely located teams where collaboration is challenging. The Tuscany Dashboard handles the entire gamut of design and physical data, displaying it via the web to the different designers involved, and providing valuable feedback to project managers. It allows everyone on the team, no matter where they’re located, to visualize the design in a way that makes the team completely collaborative and efficient in tracking progress on all fronts. The same model also applies to communications between ASIC vendors and their customers.”
Clearly, Mueller is jazzed about the company. I asked what drew him to Tuscany, given that prior to this involvement he was “retired.”
Mueller said, “They had an extremely talented and committed founding team, which had already survived through the toughest two years in EDA history. They asked me to look at their website and messaging, and I found it suggested a traditional Italian countryside ambiance, relaxing on a patio with a glass of wine overlooking a vineyard. As appealing as that may be, it did not communicate to potential customers the hard-driving nature of the Tuscany technology and team. So, as I came onboard to drive the business end, we worked to modify the image.
“The engineers at Tuscany are providing something that’s really valuable – and yes, it does add some level of Tuscan ‘relaxation’ to the process in that it helps handle the complexity of remotely located teams doing huge designs. More importantly, however, designs will move more quickly and predictably to completion using our technology. That’s the message I wanted to help the company emphasize and develop.
“An additional part of the ‘rebranding’ that we‘ve undertaken since I started in June has been to shorten and modify the original name of the flagship product, Tegula, which refers to an Italian roofing tile. The product is now called Tego, invoking a reference to the same interoperable, structured design associated with Lego building blocks that we’re all familiar with.
“We want people to focus on the fact that our technology helps fit together the disparate parts of a design, and the disparate designers on a team, as easily as Legos fit together. As geometries continue to shrink and design complexity goes off the charts, a structured design methodology becomes more and more critical.”
He added, laughing, “Of course, if design teams and our customers are more relaxed as a result of using our technology, then Tuscany is exactly the right cultural reference.”
A passion for startups …
Keith Mueller has a passion for startups and the out-of-the box creativity such enterprises entail, with a track record of success that includes in reverse chronological order:
* Apache, where he was Employee #8, and served as VP of Worldwide Sales and Marketing, taking the company from $0 to over $20 million in sales during his 4-year tenure prior to his sabbatical in 2006. Mueller noted the company continues to grow, and is primed for additional success as the global and EDA economies improve.
* Silicon Perspective, where Mueller was also Employee #8, also served as VP of Worldwide Sales, and also took the company from $0 to $20 million, prior to its acquisition by Cadence in 2001.
* Anagram, where he was Employee #4, VP of Worldwide Sales, and helped build the company up to the point of its acquisition by Avanti.
* Quickturn, where he was Employee #45, and part of the team that took the company public prior to its acquisition by Cadence.
* Silicon Compilers, back in the late 1980’s, where he started in marketing and developed the foundry relationships that helped customers to compare different process option in terms of die size, power, and speed using the tool called Genesil. Mueller later made the career change into sales, selling that product which allowed designers to work at the micro-architectural level using ALU’s barrel shifters, register files, etc. Silicon Compilers eventually merged with Silicon Design Labs to becomes Silicon Compiler Systems, and was later acquired by Mentor Graphics in 1990.
Mueller said Silicon Compilers was a company ahead of its time, and added, “It’s interesting to me that in many ways, Tuscany is a combination of ideas from both Silicon Compilers and Silicon Perspective.”
Clearly, what Mueller doesn’t know about the dynamics of startups in EDA probably isn’t worth knowing. That’s also the case when it comes to mergers and acquisitions: “I’ve been through 14 M&As in my EDA career. Some have gone more smoothly, or been more successful, than others. but all of them have been memorable. And, all of them have taught me something about the business of EDA, and about myself.”
In the thick of things …
I asked Mueller if rather than heading up Tuscany, shouldn’t he be teaching Startups 101 at a business school at this point his career. He laughed and said, “Maybe, but I’m not done doing startups. I’m not ready to stand on the sidelines and teach others to do it. It’s more fun to be in the thick of the action.
“When I first came out of school, I was an IC designer. That technical background and experience helped me to better understand and relate to the complex job that my customers have to deal with. This background also was of great benefit in my EDA marketing and sales career.
“Now, combining my sales, marketing, and engineering experience, together with the customer input I solicit during my due-diligence process, helps me greatly in identifying the right startup opportunities and team. Certainly there’s an element of luck of being in the right place at the right time, but the nice thing about a startup is that you can truly impact the outcome with a lot of hard work and a willingness to participate in all phases of the business. It’s extremely stimulating.
“After I left Apache, I took 3 years off and had a number of great experiences such as traveling extensively in Africa and New Zealand, getting my instrument pilot rating, and spending quality time with my family. I also did a major remodel of our house during the first 15 months, which was a much more stressful experience than most startups I’ve been in. But, when I saw Tuscany and went through the due diligence of examining the company, it became obvious to me that the underlying quality of the company, the team, and the technology was profound. In my experience, a startup only succeeds when you have strong momentum on all fronts simultaneously – in R&D, in sales, and in management – and we have that at Tuscany.
“We’re competing head-to-head with internally developed solutions and filling a major gap in the offerings of other EDA suppliers. I’m positive that what Tuscany is offering, and will offer in the near future, provides a great deal of value to our customers, and will therefore drive our success and return to investors. This is exactly why I decided to come back out of ‘retirement’ – to enjoy the challenge and stimulation of working at an EDA startup once again.
“Now we’re a team, and I couldn’t be luckier. These folks have been working to carefully preserve their capital as they’ve developed the technology and persevered through this tough economy. I thoroughly believe the company’s now at a point in its trajectory where we can grow, establish and expand on the message, and take the products and the presence to the next phase of growth. It doesn’t get any better than this in terms of timing to be jumping in with both feet.”
Mueller laughed again and said, “Call me crazy, but I’m excited to be back in EDA!”
Friday, September 4th, 2009
If you missed the opening keynote at Mentor Graphics’ globe-trotting EDA Tech Forum at the Santa Clara Convention Center on September 3rd, you missed hearing from the future of the automobile industry.
Ian Wright, Founder & CEO of Wrightspeed, spoke for an hour on Thursday morning and articulated his vision for the future of electric-drive vehicles. He started by announcing a revolution is at hand, but then spent the next 60 minutes explaining why the migration from combustion-engine vehicles to electric-drive vehicles will take place over the next decade at an evolutionary, not revolutionary pace.
Additionally, and despite the sexiness of the one-off Wrightspeed X1 roadster on full display in the center of the EDA Tech Forum Exhibition Hall, Wright also carefully explained during his keynote address that it’s not the family car, or light-weight, high-performance race cars, where the immediate implementation of electric-drive will succeed.
Instead, Wright said that over the next decade, big old kludgy bread trucks, FedEx delivery vans, and the ubiquitous Number 1 vehicle sold in America, the Ford F1-150, as well as many other types of pick-up trucks, will convert to electric-drive at such a pace that by 2019, upwards of 50 percent of the entire class of “high annual fuel consumption” vehicles and fleets will have left their combustion engines behind.
Per Wright, the tipping point that guarantees the requisite 3-year payback for the cost of converting to electric-drive in big, heavily used, gas-guzzling vehicles is only $3 to $4 per gallon of gasoline. For the “family car”, however, you’ve got to see gas skyrocket to $17 a gallon to defend the economics, and 3-year payback, of transitioning from combustion engine to electric drive.
Hence, fleets of trucks and delivery vans may be distinctly less sexy than roadsters, and designed to meet a far different set of performance metrics, but Ian Wright maintained that it’s these types of work-horse vehicles which will be at the front line of the massive “revolutionary” change to electric drive.
If that’s the case, why does Ian Wright need to show up at engineering conferences like EDA Tech Forum with his speedy X1 in tow? Who cares that his handcrafted vehicle, which looks like a cross between a cantilevered bridge and an old-time go cart, can beat a top-of-the-line Porsche or Ferrari in a classic 0-to-60 mph dust-up?
Well, if you were running a conference and wanted to raffle off a joy ride in a proof-of-concept vehicle, do you think anybody would interested in winning a trip in a FedEx Van? Right. Wright.
Of course, you’d want to raffle off the chance to experience 0-to-60 in 2.9 seconds in a way-cool go-cart, equipped only with shoulder restraints but no visible roll bar.
Also, if your keynote speaker said he was developing “scalable digital drive system platforms”, do you think hundreds of people would show up to hear his talk? Right. Wright.
If, however, your speaker used to work at sexy Tesla Motors, and now heads up Wrightspeed – invoking his own name, and Wilbur & Orville all at the same time – wouldn’t you want Mr. Right to show up with something that made everybody think about Tesla, even if only subliminally?
Okay, enough of all of that. So, what else is Ian Wright’s crystal ball telling him? Per his talk:
* Electric drive is completely, totally going to be the future of the vehicle industry, with the motor driving the wheels directly. Forget the transmission (and, ergo, forget Tesla’s 6831 Li-batteries strapped to same), because sleek, brushless motors (with no more moving parts than 3 ball bearings and a couple of gears) are going to be doing the job, placed right at the wheel.
* Cars are no longer about the magic of superb mechanical design. Going forward, cars will be about the magic of battery systems, electric motors and drive electronics, generator control systems, vehicle dynamic controls, the UI, and software control planes. In other words, cars are going to be about the uber-business model of Wrightspeed. Forget Detroit. The buzz has moved to Silicon Valley.
* Although the jury’s still out as far as the battery is concerned, Lithium chemistry that will win in the end, as opposed to a host of other contenders including nickel-metal hydride (NiMH). Plus, Lithium’s not an environmental toxin. QED.
* The reason the time frame for conversion to electric-drive should be revolutionary, but is actually evolutionary, is because the national infrastructure and the grid are just not ready. Wright said you can head out from California today, destined for New York, in your combustion-enabled vehicle without worrying about finding gas along the way. If, however, you set out on that same trip in an eVehicle, you should be sure to pack some worry along with the luggage. Where are you going to recharge your eV along the way?
* No matter how much Spectrum and Science Friday may wax poetic over fuel cells in trains or streetcars, the future’s about batteries, batteries, and more batteries. There’s no distribution system in place for Hydrogen, nor any plans to put one there, according to Wright. So, if you don’t like batteries, you’d be far better off putting your money on natural gas, a widely available domestic product, than to go with the sci-fi enthusiasts’ fuel-cell vision of the future.
* Ditto for photovoltaics. There’s just not enough real estate on top of a car, bus, or train to capture the photons needed, especially in places that don’t enjoy the sunny clime of Southern California or Arizona. Of course, if somebody were to produce some kind of breakthrough with the photovoltaic thing, that might change. But, remember: Wright’s talk was all about evolution, not revolution.
So, is it disappointing that the Wright Speed for this evolution is one that recognizes the realities of the engineering challenges and the national infrastructure? Absolutely not, because Ian Wright is the future of the automobile industry. And change takes time.
It also takes money.
Have you got a spare $40 million dollars sitting around? If so, give Ian a call, because you should be hitching your star, your investment dollars, your design expertise, and your vision of the transition to electric-drive to his organization, his system-level design scheme, and his X1 roadster and vision.
The future of the automobile industry is upon us, and Wrightspeed’s going to get there – wherever and whatever there is – a lot faster than anybody else.
In about 2.9 seconds.