New materials aren't free

The lead editorial (PDF file, free registration required) in this month's issue of the MRS Bulletin observes that few researchers know how much the materials they're investigating actually cost. The author makes an extremely important point. While it's true that cost is negligible for research quantities of most materials, it most certainly is not negligible for industrial quantities. In fact, cost is one of the reasons why silicon dominates both the photovoltaic and integrated circuit markets, and one of the most serious obstacles to any competing technology. Researchers who choose to ignore cost considerations are likely to be rudely surprised when industrial interest in their creations fails to materialize.

Well, maybe not

You know what they say about best laid plans... Only 10,580 words since my last word count post, which gives a total of 97,880 for the year. 150,000 is still within reach, but does require a bit more consistency.

This is not a drill, those are real dollars you're losing

What is six weeks of production worth? That's what ASMC says it will lose after a power outage at the company's Shanghai fab. Back in March, when Samsung lost power at six fabs in Giheung, South Korea, to a faulty switchboard, all six were back online within a day or so. I don't know what accounts for the difference between the two, but I would say Samsung's superior disaster preparation paid for itself.

Going along to get along in China

The current congressional hearings into Yahoo's role in the arrest of several Chinese dissidents should serve as a cautionary tale for companies doing business in China and other repressive states. On the one hand, I have a lot of sympathy for Yahoo's local employees in China: they really had no reasonable choice but to comply with Chinese law, no matter how repressive that law may be. Yahoo isn't paying them enough to become prisoners of conscience themselves.

I have no sympathy whatsoever for Yahoo's executives, however. China's appalling human rights record and vicious suppression of dissent are well-documented. They should have known that sooner or later someone would use their service for the "free exchange" of information that the Chinese government wouldn't like, and that they'd be stuck between American principles and Chinese laws. The backlash they are now experiencing is a completely predictable consequence of their approach to the Chinese market.

It's easy to say that any company doing business in China has to comply with the same laws, and any company avoiding China on human rights grounds is likely to place itself at a competitive disadvantage. In reality, Human Rights Watch explains, there are laws, and there's a gray area of unwritten understandings. For instance, some companies proactively censor material that they think might be objectionable, while others block only material that has been specifically banned. Some servers are physically located in China, under the jurisdiction of Chinese courts, while some are not. Some companies respond to criticism by changing their policies, some by spinning out Chinese subsidiaries that they cannot control. It's up to customers and shareholders to differentiate between those that try to maintain open information principles under very difficult circumstances, and those that toss principles under the bus as soon as it's expedient.

Don't scrap that wafer

IC manufacturers throw away about three million prime wafers a year. Solar cell manufacturers are getting squeezed by a silicon shortage.

If that sounds like a business opportunity waiting for a product, you're right. IBM has figured out a way to de-pattern scrap wafers, cleansing them of IP so that they can be recycled.

Now, reclaimed wafers are nothing new. I suspect the difference is a better way to get rid of the IP, but I'm waiting to hear more from IBM.