By David Lammers

Bob Johnstone, a technology journalist and author of five books, has brought his story-telling skills to the immensely important role that solar energy can play in creating a cleaner world. His latest book, Switching to Solar, (Prometheus Books, 2011) is as well-researched as his previous volume “Brilliant!,” on the LED industry, which I also heartily recommend.

I should reveal my assumptions. First, I respect the scientific consensus that global warming is real, that the main cause is the steady addition of carbon dioxide into the atmosphere, and that we have to reduce the burning of coal, oil, and gas if we want to prevent global catastrophe.

Second, I am biased towards Mr. Johnstone, who as a magazine journalist lived in Tokyo at the same time I did, during the heyday of the Japanese electronics industry. A Scot by birth, he moved with his family from Tokyo to sunny Australia more than a decade ago to begin his book-writing career.

Johnstone excels as an historian of technology, weaving the history of a technology into an account of its current challenges and future prospects. For example, I read an entire biography of Thomas Edison 10 years ago but don’t recall the great inventor decrying the use of coal and oil, likening the practice to “tenant farmers chopping down the fence around our house for fuel,” and referring to the sun as “What a source of power! I hope we don’t have to wait until oil and coal run out before we tackle that.” Quotes like that are the reason Johnstone’s books are such a pleasure.

Johnstone takes us through a brief history of solar, beginning with Albert Einstein’s 1920s discovery of the photovoltaic effect (for which he won the Nobel Prize), and including Daryl Chapin’s creation of an early solar cell at Bell Labs in 1953.

Johnstone peppers his books with great anecdotes; he has a knack for weaving mini-biographies into his books, describing how strong, colorful personalities are able to impact solar’s role in our energy policies.What could be tedious reading in another author’s hand comes to life as Johnstone melds human-interest stories about California’s hectic energy scene into Switching to Solar.

Much of “Switching to Solar” (which has the sub-title What We Can Learn from Germany’s Success in Harnessing Clean Energy) is about Germany’s implementation of a feed-in tariff, which guarantees a premium rate for a period of 20 years for all electricity generated and fed into the grid by owners of solar systems. “The whole idea is that solar generators should be fully compensated for the cost of their systems, plus make a return on investment, just like utilities do with their power plants,” Johnstone said.

The goal of the Europeans is to source 12 percent of their energy from solar by 2020.  In the south of Germany, solar averages about three percent of its electricity, increasing to 20 percent or more on a hot afternoon.

What lessons can others learn from Germany’s success with solar? “Keep it simple,” Johnstone advises, with an easy-to-understand policy so that people who want to invest in solar know what they’re signing up for.

Johnstone spends several years researching each of his books. For Switching to Solar he interviewed more than 80 people, including as many as four long sessions with several of the early pioneers of solar energy. Thanks to those interviews, the book is a lively read, all the more so because its topic is truly essential to saving the planet as we know it.

By David Lammers

A few years ago, access to capital amid a financial meltdown was the big challenge for chip manufacturers. Now, it is finding the right people. Looming on the horizon are rising costs for energy, fab managers from Freescale, Samsung, and Spansion said at a May 11 SEMI Austin breakfast forum.

Michael Raiford

Michael Raiford, senior director of the manufacturing and systems division at Samsung Austin Semiconductor, gave a close look inside the culture at one of the largest – and fastest growing — fab complexes in North America. Samsung Austin will install its last tool tomorrow (Friday) for the $3.6B logic expansion to its Fab 2, where SoCs for smart phones and tablet computers will be made. Until now, the Austin fab has concentrated largely on memory production, including high volumes of a single product: 32-Gbit NAND, eight of which are required for every advanced smart phone.

“Our biggest problem? People,” Raiford told about 115 people at the SEMI Austin event. “We’ve gone from 1,000 to 2,000 people. We are still trying to hire more people, and we can’t find enough of them.”

Half of the people now working at the Austin fab have been hired in the past year, and integrating these new-hires is the biggest challenge facing the existing staff. And Samsung is competing with other companies to hire experienced technicians who can keep the fab production equipment in good working order, Raiford said.

One good source has been the military. “We have hired a ton of ex-military” – about 220 – “going all around the country to find them,” Raiford said. After mentioning Samsung’s set of rules for how to operate inside a Samsung fab, Raiford drew a laugh when he said “people in the military are used to following rules, so they fit in well with Samsung. The best have been the ex-submariners — they are trained to not make any mistakes down there.”

The employee spectrum is changing. A few years ago, Samsung Austin employed 450 operators, which has dropped to 120 as automation at the 300-mm fab has taken hold. However, total employment has increased sharply as more technicians and engineers have been employed. Samsung wants to train its maintenance technicians well, so they can be more independent. They must be able to quickly fix boards and systems without relying on vendors.

Not everyone is a good fit with the Samsung culture. Samsung “wants to be no. 1 — that’s our goal. We are a Korean company, with a different culture. Not everyone finds that appealing. We are in a constant crisis mode, which some people don’t find very pleasing. As our chairman explained to Bloomberg last year, we want to use a little bit less equipment, with shorter timelines, and get more output than we think we are capable of doing. We are in a constant crisis mode, to always be better. Not everyone thrives on that,” Raiford said.

Originally trained in computer science, Raiford has played a key role in helping to automate the business process management (BPM) system, which guides people to better decision-making. “Having so many inexperienced people forces changes. Half of our people have been at our company a year or less. When we say: ‘Let’s go back to basics,’ that has a new meaning when half of them might not know what the basics are.”

The BPM system is aimed at preventing problems. “We want to use automation to get people to make the right decisions, but still allow for creativity. Our challenge is: How do you have an environment with a lot of structure in it, and still allow creativity?”

Gary Dawson

Gary Dawson, director of operations at Spansion’s Fab 25, said the maintenance technicians at Spansion are learning to repair boards and do other tasks that might have required a service call from an equipment vendor. “We are cutting back on outside support, with fewer service contracts, as part of our program to avoid long downs. We are training our maintenance department, and making sure they have the parts needed to make repairs in a more flexible manner.” The skill level of the maintenance crew at Spansion is a key element in the emergence from Chapter 11 by the embedded flash manufacturer, which had $1.2 billion in revenues last year.

Chris Magnella, manager of Freescale Semiconductor’s Oak Hill Technology (OHT) fab, warned that the city of Austin could put fabs out of business if it follows through on promised rate increases for electric power. “The cost of power is a major issue for us. Any company which uses power in a large way in this city needs to understand what is going on,” he said. Austin owns its electric utility, and Austin Power is moving to adopt more solar, wind, and other forms of energy to reduce carbon emissions, but at higher utility rates.

Chris Magnella

Magnella said the three major fab operators in Austin – Freescale, Samsung, and Spansion – should “cooperate with the City of Austin on the cost of power. That has the potential to get very expensive.”

The Oak Hill fab is moving rapidly to take on some of the MEMS production formerly done at the now-closed Freescale Sendai fab. After the March 11 earthquake, Freescale was forced to accelerate the transition of its automotive-use MEMS from Sendai to Austin. He said the premature closure of the Sendai fab was unfortunate, but quoted race car driver Mario Andretti as saying: “If everything seems under control, you are not going fast enough.”

By David Lammers

As I listened to the Intel 22nm presentation on the benefits of their tri-gate transistor, I thought back to the day when the journalist Lew Simons heard that he had won his second Pulitzer Prize, for stories about political corruption in the Philippines.

“Yeah, my editor just called me,” Simons said at the Tokyo press club, as champagne glasses were raised in congratulations. “He offered his regards, and then said: ‘Those stories were last year, Lew. What did you do for me yesterday?’”

That tradition in old-school journalism, of providing clean copy while keeping the excuses and B.S. to a minimum (editors used to have pressroom foremen yelling at them, saying it was time to roll the presses) is one that permeates every industry.

There was quite a bit of that “What did you do for me yesterday?” expressed at the Intel briefing on May 4th. Intel senior fellow Mark Bohr made an attempt to steer the discussion towards device physics, subthreshold slopes, and the things that people like me want to hear about. And to be sure, even the New York Times made a decent effort to report that Intel had made a historic jump by shifting to a vertical transistor. (It should have been on page one).

However, most of the questions from the reporters at the Intel briefing had to do with the competition with ARM-based cores, the prospects for the Atom processor, Intel’s competitiveness in the mobile space, and so on. Dadi Perlmutter, general manager of the Intel Architecture operation, gave an interesting answer to one reporter’s question. After noting that Intel has much work to do on the software and “ecosystem” surrounding the Atom, he said, “We want to accelerate the Atom roadmap, and close the gap at 22nm. At one point we want to ramp all of our products at the same time.”

An important point, one not to be overlooked or undervalued, is that Intel has found a way to stay on Moore’s Law. I seem to run into lots of curmudgeonly old guys who believe transistor scaling either has already ended, or will end soon, or won’t be affordable, or will end up in the hands of people in Korea and Taiwan.

Source: Intel Corp.

With its P1270 rollout, Intel showed that the 22nm tri-gate succeeds not only on density scaling, impressive (but not enough) at these bleeding-edge dimensions. By enabling a technology that apparently operates well at 0.8 V, Intel is saying that it can reduce power consumption by 10x at a given performance level, or boost performance by 37 percent at a given active power.

P = CV²f. Capacitance times the voltage squared times frequency equals power: that quadratic on the voltage factor makes it a powerful knob to turn.

In a sense, Bohr’s emphasis on 0.8 V operation is a rebuttal, a competitive thrust. The SOI Consortium said in early February that it had tested out an ARM-based SoC, fabbed on 32nm SOI at IBM, and had found good operation at 0.8 V. Of course, there were no specifications, no data on power consumption, either static or dynamic.

Intel also was predictably vague on its raw power consumption numbers. Perhaps more precise information will be published at the IEDM 2011 in Washington, D.C.

We live in a society where no one gets to rest on his or her laurels, not with the world population heading towards 10 billion and the polar ice caps melting.

For just one moment, however, perhaps we could feel a sense of relief or just plain head-shaking wonder that transistors can be made so small. Intel’s vertical transistor appears to deliver the goods, and appears to be manufacturable – by at least one company.

Now, even as we raise a glass of virtual bubbly towards Hillsboro, let the battle continue between the finFET camp and the ultra-thin-body silicon-on-insulator (UTB-SOI) proponents.

It is a great story. Let’s get to work on it.

Source: Intel Corp.