Former Intel CEO Gordon Moore, creator of Moore’s Law, in his cubicle at Intel.
In 1965, Intel co-founder Gordon Moore observed that over the history of computing hardware, the number of transistors on integrated circuits doubles approximately every two years.
Moore always said it was just a hypothesis, but somehow, the semiconductor industry has managed to meet or exceed Moore’s Law ever since. It’s the only industry that has ever innovated at that kind of rate.
Moore’s Law is only 11 words long, but it’s had a profound effect on the world. Some treat Moore’s Law as if it were some kind of scientific rule – but it’s not a law of physics, just an uncannily accurate observation about what engineers can achieve.
Just as importantly, companies don’t follow Moore’s Law out of the goodness of their hearts. They do it because those that keep innovating grow and prosper, while those that don’t fade away.
So the semiconductor industry needs continual, major innovation in nanotechnology – and if that innovation doesn’t happen, the kind of growth we’ve seen in the last 100 years will stop. Imagine also, if rapid innovation as Moore predicted was not limited to semiconductors alone, but extended throughout industry?
Predicting the Demise of Moore’s Law
Many executives and pundits, from Google’s Eric Schmidt to Red Herring Magazine founder Tony Perkins have predicted or proclaimed the end of Moore’s Law. Perkins did it in a cover story on the death of Moore’s Law in February 2003 – ironically, just weeks before Red Herring went out of business.
But somehow, the industry just keeps producing more and more powerful chips at smaller and smaller size. But over the last two decades it’s become harder and harder – and more and more expensive – to keep doubling the processing power while shrinking the size of the circuits.
Why? Because you can only get to the next level by making things smaller and taking advantage of the innovations and increases in processing power and speed that are made possible by getting smaller.
Nanotronics Imaging nSPEC System
You can’t build it if you can’t see it. If you can’t see it, how do you build the circuits themselves, manage the manufacturing plants and assembly lines, handle quality control, and all the other things that you need to continue to innovate by making things smaller? One of the problems that the fabrication industry has struggled with is finding new ways to inspect the ever-smaller and ever more complex circuits in the numbers required for efficient quality control. Many advances in microscopy such as scanning probe microscopy and infrared microscopy have been developed to help keep up with the demands of the semiconductor fabrication industry, but due to the cost and low throughput, they have primarily helped with design – not manufacturing. Unfortunately, most of the advances in microscopy relied on methods that just aren’t suitable for large-scale qualify control. They show excellent images of a single thing, but the time and skilled man-hours required to set up each image is simply not workable in a manufacturing environment and the costs are prohibitive for smaller companies.So how is the industry going to keep up? According to Nanotronics Imaging co-founder Matthew Putman, by adopting disruptive technology that enables workers to be able to see a lot of very small things, very quickly, at a very affordable price.
More Than Meets the Eye
“I worked in a family company, an instruments maker called Tech Pro. In the late 1990’s, I was frustrated because we lost customers because they needed to be able to see a lot of very small things, very quickly, at a very affordable price. I couldn’t solve that problem for them – in fact, no one could solve that problem for them, or for us,” Putman says.
So he went back to college for a PhD in physics, and invested more than a million dollars of the money earned from the sale of Tech Pro into developing a hybrid system that combined best-of-class optical microscopes with algorithms he developed for the software that powers a new kind of optical microscopy solution for manufacturers.
“It’s a disruptive technology because it doesn’t replace anything that existed before,” Putman says. “We’ve developed a system that removes a barrier to innovation by solving a problem that was unsolvable until several things happened at once. The processing power of computers had to reach a certain point – and someone had to do the mathematics required to create the algorithms needed to handle the processing and imaging required to deliver nanoscale images in real-time – and the resulting software had to be integrated with top-of-the-line microscopes.”
Particle Defect Map
The result is now in use by more than 20 customers who are using the Nanotronics Imaging nSPEC® system to automate the inspection process. “Instead of sampling the production line’s chips for defects, they can now inspect every single one for defects or other features of interest,” Putman says.
Customers are using the optical microscopy systems for silicon carbide and Gallium Nitride epi wafers. “What they can achieve is faster inspection, with detailed reporting and mapping. It’s fully automated, so the operator can set it and walk away, yet when a report identifies a potential problem, it’s easy to take a closer look,” he says.
Manufacturers who need the sub-nanometer imaging possible with atomic force microscopes (AFMs) can attach an optional AFM tip directly to the nSPEC system. “Our system offers 50 nanometer resolution now, and we’re moving toward 100 nanometer resolution,” Putman adds.
Is Moore’s Law Doomed Anyway?
The advances that Nanotronics Imaging is bringing to the wafer fab process are critical in keeping up with the challenges that manufacturer’s face. But, long term, Moore’s Law may be unsustainable.
Processing speeds have increased over a million-fold since Gordon Moore described his law in 1965, but researchers are starting to realize that without some sort of intervention the technology will hit a few logistical and physical limits around 2020. Among the problems that experts think may slow innovation are a shortage of R&D funding as well as a global lack of semiconductor engineers.
The Information Technology and Innovation Foundation put together a panel discussion in 2013 that brought together experts from academia and business to look at the future of Moore’s Law. One fact that the panel found alarming is the fact that every second, 640 Terabits of data are traveling through the Internet and that number is doubling and quadrupling every year. By 2016, consumers will be accessing 16 times as much video as they are viewing online right now – and that means a dramatic demand for more powerful processors, something the panel says isn’t possible even if Moore’s law is sustained.
Dr. Putman isn’t that pessimistic. “It’s easy to forget just how much of our future hasn’t yet been decided because we don’t have any idea of what exciting discoveries will be made in labs all over the world. Very talented people are exploring new chip architecture, spectral analysis, reconstruction techniques, and illumination systems that will create opportunities that are now only being imagined.
“I’m looking forward to experimenting with our partners, and sharing ideas with universities, clients, and partners. I think we’re at the beginning of a new age of innovation, not the end of an era.”
Photo credits: The photograph of former Intel CEO Gordon Moore was offered on Flickr under a Creative Commons License. The photograph of the Nanotronics Imaging nSPEC system and the particle defect map are the property of Nanotronics Imaging.